File size: 161,514 Bytes
05922fb |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856 2857 2858 2859 2860 2861 2862 2863 2864 2865 2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960 2961 2962 2963 2964 2965 2966 2967 2968 2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 3083 3084 3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111 3112 3113 3114 3115 3116 3117 3118 3119 3120 3121 3122 3123 3124 3125 3126 3127 3128 3129 3130 3131 3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475 3476 3477 3478 3479 |
# Modified version of nltk framenet
# Natural Language Toolkit: Framenet Corpus Reader
#
# Copyright (C) 2001-2020 NLTK Project
# Authors: Chuck Wooters <[email protected]>,
# Nathan Schneider <[email protected]>
# URL: <http://nltk.org/>
# For license information, see LICENSE.TXT
"""
Corpus reader for the FrameNet 1.7 lexicon and corpus.
"""
import os
import re
import textwrap
import itertools
import sys
import types
from collections import defaultdict, OrderedDict
from operator import itemgetter
from itertools import zip_longest
from pprint import pprint
from nltk.corpus.reader import XMLCorpusReader, XMLCorpusView
from nltk.corpus.util import LazyCorpusLoader
from nltk.util import LazyConcatenation, LazyMap, LazyIteratorList
__docformat__ = "epytext en"
def mimic_wrap(lines, wrap_at=65, **kwargs):
"""
Wrap the first of 'lines' with textwrap and the remaining lines at exactly the same
positions as the first.
"""
l0 = textwrap.fill(lines[0], wrap_at, drop_whitespace=False).split("\n")
yield l0
def _(line):
il0 = 0
while line and il0 < len(l0) - 1:
yield line[: len(l0[il0])]
line = line[len(l0[il0]) :]
il0 += 1
if line: # Remaining stuff on this line past the end of the mimicked line.
# So just textwrap this line.
for ln in textwrap.fill(line, wrap_at, drop_whitespace=False).split("\n"):
yield ln
for l in lines[1:]:
yield list(_(l))
def _pretty_longstring(defstr, prefix="", wrap_at=65):
"""
Helper function for pretty-printing a long string.
:param defstr: The string to be printed.
:type defstr: str
:return: A nicely formated string representation of the long string.
:rtype: str
"""
outstr = ""
for line in textwrap.fill(defstr, wrap_at).split("\n"):
outstr += prefix + line + "\n"
return outstr
def _pretty_any(obj):
"""
Helper function for pretty-printing any AttrDict object.
:param obj: The obj to be printed.
:type obj: AttrDict
:return: A nicely formated string representation of the AttrDict object.
:rtype: str
"""
outstr = ""
for k in obj:
if isinstance(obj[k], str) and len(obj[k]) > 65:
outstr += "[{0}]\n".format(k)
outstr += "{0}".format(_pretty_longstring(obj[k], prefix=" "))
outstr += "\n"
else:
outstr += "[{0}] {1}\n".format(k, obj[k])
return outstr
def _pretty_semtype(st):
"""
Helper function for pretty-printing a semantic type.
:param st: The semantic type to be printed.
:type st: AttrDict
:return: A nicely formated string representation of the semantic type.
:rtype: str
"""
semkeys = st.keys()
if len(semkeys) == 1:
return "<None>"
outstr = ""
outstr += "semantic type ({0.ID}): {0.name}\n".format(st)
if "abbrev" in semkeys:
outstr += "[abbrev] {0}\n".format(st.abbrev)
if "definition" in semkeys:
outstr += "[definition]\n"
outstr += _pretty_longstring(st.definition, " ")
outstr += "[rootType] {0}({1})\n".format(st.rootType.name, st.rootType.ID)
if st.superType is None:
outstr += "[superType] <None>\n"
else:
outstr += "[superType] {0}({1})\n".format(st.superType.name, st.superType.ID)
outstr += "[subTypes] {0} subtypes\n".format(len(st.subTypes))
outstr += (
" "
+ ", ".join("{0}({1})".format(x.name, x.ID) for x in st.subTypes)
+ "\n" * (len(st.subTypes) > 0)
)
return outstr
def _pretty_frame_relation_type(freltyp):
"""
Helper function for pretty-printing a frame relation type.
:param freltyp: The frame relation type to be printed.
:type freltyp: AttrDict
:return: A nicely formated string representation of the frame relation type.
:rtype: str
"""
outstr = "<frame relation type ({0.ID}): {0.superFrameName} -- {0.name} -> {0.subFrameName}>".format(
freltyp
)
return outstr
def _pretty_frame_relation(frel):
"""
Helper function for pretty-printing a frame relation.
:param frel: The frame relation to be printed.
:type frel: AttrDict
:return: A nicely formated string representation of the frame relation.
:rtype: str
"""
outstr = "<{0.type.superFrameName}={0.superFrameName} -- {0.type.name} -> {0.type.subFrameName}={0.subFrameName}>".format(
frel
)
return outstr
def _pretty_fe_relation(ferel):
"""
Helper function for pretty-printing an FE relation.
:param ferel: The FE relation to be printed.
:type ferel: AttrDict
:return: A nicely formated string representation of the FE relation.
:rtype: str
"""
outstr = "<{0.type.superFrameName}={0.frameRelation.superFrameName}.{0.superFEName} -- {0.type.name} -> {0.type.subFrameName}={0.frameRelation.subFrameName}.{0.subFEName}>".format(
ferel
)
return outstr
def _pretty_lu(lu):
"""
Helper function for pretty-printing a lexical unit.
:param lu: The lu to be printed.
:type lu: AttrDict
:return: A nicely formated string representation of the lexical unit.
:rtype: str
"""
lukeys = lu.keys()
outstr = ""
outstr += "lexical unit ({0.ID}): {0.name}\n\n".format(lu)
if "definition" in lukeys:
outstr += "[definition]\n"
outstr += _pretty_longstring(lu.definition, " ")
if "frame" in lukeys:
outstr += "\n[frame] {0}({1})\n".format(lu.frame.name, lu.frame.ID)
if "incorporatedFE" in lukeys:
outstr += "\n[incorporatedFE] {0}\n".format(lu.incorporatedFE)
if "POS" in lukeys:
outstr += "\n[POS] {0}\n".format(lu.POS)
if "status" in lukeys:
outstr += "\n[status] {0}\n".format(lu.status)
if "totalAnnotated" in lukeys:
outstr += "\n[totalAnnotated] {0} annotated examples\n".format(
lu.totalAnnotated
)
if "lexemes" in lukeys:
outstr += "\n[lexemes] {0}\n".format(
" ".join("{0}/{1}".format(lex.name, lex.POS) for lex in lu.lexemes)
)
if "semTypes" in lukeys:
outstr += "\n[semTypes] {0} semantic types\n".format(len(lu.semTypes))
outstr += (
" " * (len(lu.semTypes) > 0)
+ ", ".join("{0}({1})".format(x.name, x.ID) for x in lu.semTypes)
+ "\n" * (len(lu.semTypes) > 0)
)
if "URL" in lukeys:
outstr += "\n[URL] {0}\n".format(lu.URL)
if "subCorpus" in lukeys:
subc = [x.name for x in lu.subCorpus]
outstr += "\n[subCorpus] {0} subcorpora\n".format(len(lu.subCorpus))
for line in textwrap.fill(", ".join(sorted(subc)), 60).split("\n"):
outstr += " {0}\n".format(line)
if "exemplars" in lukeys:
outstr += "\n[exemplars] {0} sentences across all subcorpora\n".format(
len(lu.exemplars)
)
return outstr
def _pretty_exemplars(exemplars, lu):
"""
Helper function for pretty-printing a list of exemplar sentences for a lexical unit.
:param sent: The list of exemplar sentences to be printed.
:type sent: list(AttrDict)
:return: An index of the text of the exemplar sentences.
:rtype: str
"""
outstr = ""
outstr += "exemplar sentences for {0.name} in {0.frame.name}:\n\n".format(lu)
for i, sent in enumerate(exemplars):
outstr += "[{0}] {1}\n".format(i, sent.text)
outstr += "\n"
return outstr
def _pretty_fulltext_sentences(sents):
"""
Helper function for pretty-printing a list of annotated sentences for a full-text document.
:param sent: The list of sentences to be printed.
:type sent: list(AttrDict)
:return: An index of the text of the sentences.
:rtype: str
"""
outstr = ""
outstr += "full-text document ({0.ID}) {0.name}:\n\n".format(sents)
outstr += "[corpid] {0.corpid}\n[corpname] {0.corpname}\n[description] {0.description}\n[URL] {0.URL}\n\n".format(
sents
)
outstr += "[sentence]\n".format(sents)
for i, sent in enumerate(sents.sentence):
outstr += "[{0}] {1}\n".format(i, sent.text)
outstr += "\n"
return outstr
def _pretty_fulltext_sentence(sent):
"""
Helper function for pretty-printing an annotated sentence from a full-text document.
:param sent: The sentence to be printed.
:type sent: list(AttrDict)
:return: The text of the sentence with annotation set indices on frame targets.
:rtype: str
"""
outstr = ""
outstr += "full-text sentence ({0.ID}) in {1}:\n\n".format(
sent, sent.doc.get("name", sent.doc.description)
)
outstr += "\n[POS] {0} tags\n".format(len(sent.POS))
outstr += "\n[POS_tagset] {0}\n\n".format(sent.POS_tagset)
outstr += "[text] + [annotationSet]\n\n"
outstr += sent._ascii() # -> _annotation_ascii()
outstr += "\n"
return outstr
def _pretty_pos(aset):
"""
Helper function for pretty-printing a sentence with its POS tags.
:param aset: The POS annotation set of the sentence to be printed.
:type sent: list(AttrDict)
:return: The text of the sentence and its POS tags.
:rtype: str
"""
outstr = ""
outstr += "POS annotation set ({0.ID}) {0.POS_tagset} in sentence {0.sent.ID}:\n\n".format(
aset
)
# list the target spans and their associated aset index
overt = sorted(aset.POS)
sent = aset.sent
s0 = sent.text
s1 = ""
s2 = ""
i = 0
adjust = 0
for j, k, lbl in overt:
assert j >= i, ("Overlapping targets?", (j, k, lbl))
s1 += " " * (j - i) + "-" * (k - j)
if len(lbl) > (k - j):
# add space in the sentence to make room for the annotation index
amt = len(lbl) - (k - j)
s0 = (
s0[: k + adjust] + "~" * amt + s0[k + adjust :]
) # '~' to prevent line wrapping
s1 = s1[: k + adjust] + " " * amt + s1[k + adjust :]
adjust += amt
s2 += " " * (j - i) + lbl.ljust(k - j)
i = k
long_lines = [s0, s1, s2]
outstr += "\n\n".join(
map("\n".join, zip_longest(*mimic_wrap(long_lines), fillvalue=" "))
).replace("~", " ")
outstr += "\n"
return outstr
def _pretty_annotation(sent, aset_level=False):
"""
Helper function for pretty-printing an exemplar sentence for a lexical unit.
:param sent: An annotation set or exemplar sentence to be printed.
:param aset_level: If True, 'sent' is actually an annotation set within a sentence.
:type sent: AttrDict
:return: A nicely formated string representation of the exemplar sentence
with its target, frame, and FE annotations.
:rtype: str
"""
sentkeys = sent.keys()
outstr = "annotation set" if aset_level else "exemplar sentence"
outstr += " ({0.ID}):\n".format(sent)
if aset_level: # TODO: any UNANN exemplars?
outstr += "\n[status] {0}\n".format(sent.status)
for k in ("corpID", "docID", "paragNo", "sentNo", "aPos"):
if k in sentkeys:
outstr += "[{0}] {1}\n".format(k, sent[k])
outstr += (
"\n[LU] ({0.ID}) {0.name} in {0.frame.name}\n".format(sent.LU)
if sent.LU
else "\n[LU] Not found!"
)
outstr += "\n[frame] ({0.ID}) {0.name}\n".format(
sent.frame
) # redundant with above, but .frame is convenient
if not aset_level:
outstr += "\n[annotationSet] {0} annotation sets\n".format(
len(sent.annotationSet)
)
outstr += "\n[POS] {0} tags\n".format(len(sent.POS))
outstr += "\n[POS_tagset] {0}\n".format(sent.POS_tagset)
outstr += "\n[GF] {0} relation{1}\n".format(
len(sent.GF), "s" if len(sent.GF) != 1 else ""
)
outstr += "\n[PT] {0} phrase{1}\n".format(
len(sent.PT), "s" if len(sent.PT) != 1 else ""
)
"""
Special Layers
--------------
The 'NER' layer contains, for some of the data, named entity labels.
The 'WSL' (word status layer) contains, for some of the data,
spans which should not in principle be considered targets (NT).
The 'Other' layer records relative clause constructions (Rel=relativizer, Ant=antecedent),
pleonastic 'it' (Null), and existential 'there' (Exist).
On occasion they are duplicated by accident (e.g., annotationSet 1467275 in lu6700.xml).
The 'Sent' layer appears to contain labels that the annotator has flagged the
sentence with for their convenience: values include
'sense1', 'sense2', 'sense3', etc.;
'Blend', 'Canonical', 'Idiom', 'Metaphor', 'Special-Sent',
'keepS', 'deleteS', 'reexamine'
(sometimes they are duplicated for no apparent reason).
The POS-specific layers may contain the following kinds of spans:
Asp (aspectual particle), Non-Asp (non-aspectual particle),
Cop (copula), Supp (support), Ctrlr (controller),
Gov (governor), X. Gov and X always cooccur.
>>> from nltk.corpus import framenet as fn
>>> def f(luRE, lyr, ignore=set()):
... for i,ex in enumerate(fn.exemplars(luRE)):
... if lyr in ex and ex[lyr] and set(zip(*ex[lyr])[2]) - ignore:
... print(i,ex[lyr])
- Verb: Asp, Non-Asp
- Noun: Cop, Supp, Ctrlr, Gov, X
- Adj: Cop, Supp, Ctrlr, Gov, X
- Prep: Cop, Supp, Ctrlr
- Adv: Ctrlr
- Scon: (none)
- Art: (none)
"""
for lyr in ("NER", "WSL", "Other", "Sent"):
if lyr in sent and sent[lyr]:
outstr += "\n[{0}] {1} entr{2}\n".format(
lyr, len(sent[lyr]), "ies" if len(sent[lyr]) != 1 else "y"
)
outstr += "\n[text] + [Target] + [FE]"
# POS-specific layers: syntactically important words that are neither the target
# nor the FEs. Include these along with the first FE layer but with '^' underlining.
for lyr in ("Verb", "Noun", "Adj", "Adv", "Prep", "Scon", "Art"):
if lyr in sent and sent[lyr]:
outstr += " + [{0}]".format(lyr)
if "FE2" in sentkeys:
outstr += " + [FE2]"
if "FE3" in sentkeys:
outstr += " + [FE3]"
outstr += "\n\n"
outstr += sent._ascii() # -> _annotation_ascii()
outstr += "\n"
return outstr
def _annotation_ascii(sent):
"""
Given a sentence or FE annotation set, construct the width-limited string showing
an ASCII visualization of the sentence's annotations, calling either
_annotation_ascii_frames() or _annotation_ascii_FEs() as appropriate.
This will be attached as a method to appropriate AttrDict instances
and called in the full pretty-printing of the instance.
"""
if sent._type == "fulltext_sentence" or (
"annotationSet" in sent and len(sent.annotationSet) > 2
):
# a full-text sentence OR sentence with multiple targets.
# (multiple targets = >2 annotation sets, because the first annotation set is POS.)
return _annotation_ascii_frames(sent)
else: # an FE annotation set, or an LU sentence with 1 target
return _annotation_ascii_FEs(sent)
def _annotation_ascii_frames(sent):
"""
ASCII string rendering of the sentence along with its targets and frame names.
Called for all full-text sentences, as well as the few LU sentences with multiple
targets (e.g., fn.lu(6412).exemplars[82] has two want.v targets).
Line-wrapped to limit the display width.
"""
# list the target spans and their associated aset index
overt = []
for a, aset in enumerate(sent.annotationSet[1:]):
for j, k in aset.Target:
indexS = "[{0}]".format(a + 1)
if aset.status == "UNANN" or aset.LU.status == "Problem":
indexS += " "
if aset.status == "UNANN":
indexS += (
"!"
) # warning indicator that there is a frame annotation but no FE annotation
if aset.LU.status == "Problem":
indexS += (
"?"
) # warning indicator that there is a missing LU definition (because the LU has Problem status)
overt.append((j, k, aset.LU.frame.name, indexS))
overt = sorted(overt)
duplicates = set()
for o, (j, k, fname, asetIndex) in enumerate(overt):
if o > 0 and j <= overt[o - 1][1]:
# multiple annotation sets on the same target
# (e.g. due to a coordination construction or multiple annotators)
if (
overt[o - 1][:2] == (j, k) and overt[o - 1][2] == fname
): # same target, same frame
# splice indices together
combinedIndex = (
overt[o - 1][3] + asetIndex
) # e.g., '[1][2]', '[1]! [2]'
combinedIndex = combinedIndex.replace(" !", "! ").replace(" ?", "? ")
overt[o - 1] = overt[o - 1][:3] + (combinedIndex,)
duplicates.add(o)
else: # different frames, same or overlapping targets
s = sent.text
for j, k, fname, asetIndex in overt:
s += "\n" + asetIndex + " " + sent.text[j:k] + " :: " + fname
s += "\n(Unable to display sentence with targets marked inline due to overlap)"
return s
for o in reversed(sorted(duplicates)):
del overt[o]
s0 = sent.text
s1 = ""
s11 = ""
s2 = ""
i = 0
adjust = 0
fAbbrevs = OrderedDict()
for j, k, fname, asetIndex in overt:
if not j >= i:
assert j >= i, (
"Overlapping targets?"
+ (
" UNANN"
if any(aset.status == "UNANN" for aset in sent.annotationSet[1:])
else ""
),
(j, k, asetIndex),
)
s1 += " " * (j - i) + "*" * (k - j)
short = fname[: k - j]
if (k - j) < len(fname):
r = 0
while short in fAbbrevs:
if fAbbrevs[short] == fname:
break
r += 1
short = fname[: k - j - 1] + str(r)
else: # short not in fAbbrevs
fAbbrevs[short] = fname
s11 += " " * (j - i) + short.ljust(k - j)
if len(asetIndex) > (k - j):
# add space in the sentence to make room for the annotation index
amt = len(asetIndex) - (k - j)
s0 = (
s0[: k + adjust] + "~" * amt + s0[k + adjust :]
) # '~' to prevent line wrapping
s1 = s1[: k + adjust] + " " * amt + s1[k + adjust :]
s11 = s11[: k + adjust] + " " * amt + s11[k + adjust :]
adjust += amt
s2 += " " * (j - i) + asetIndex.ljust(k - j)
i = k
long_lines = [s0, s1, s11, s2]
outstr = "\n\n".join(
map("\n".join, zip_longest(*mimic_wrap(long_lines), fillvalue=" "))
).replace("~", " ")
outstr += "\n"
if fAbbrevs:
outstr += " (" + ", ".join("=".join(pair) for pair in fAbbrevs.items()) + ")"
assert len(fAbbrevs) == len(dict(fAbbrevs)), "Abbreviation clash"
return outstr
def _annotation_ascii_FE_layer(overt, ni, feAbbrevs):
"""Helper for _annotation_ascii_FEs()."""
s1 = ""
s2 = ""
i = 0
for j, k, fename in overt:
s1 += " " * (j - i) + ("^" if fename.islower() else "-") * (k - j)
short = fename[: k - j]
if len(fename) > len(short):
r = 0
while short in feAbbrevs:
if feAbbrevs[short] == fename:
break
r += 1
short = fename[: k - j - 1] + str(r)
else: # short not in feAbbrevs
feAbbrevs[short] = fename
s2 += " " * (j - i) + short.ljust(k - j)
i = k
sNI = ""
if ni:
sNI += " [" + ", ".join(":".join(x) for x in sorted(ni.items())) + "]"
return [s1, s2, sNI]
def _annotation_ascii_FEs(sent):
"""
ASCII string rendering of the sentence along with a single target and its FEs.
Secondary and tertiary FE layers are included if present.
'sent' can be an FE annotation set or an LU sentence with a single target.
Line-wrapped to limit the display width.
"""
feAbbrevs = OrderedDict()
posspec = [] # POS-specific layer spans (e.g., Supp[ort], Cop[ula])
posspec_separate = False
for lyr in ("Verb", "Noun", "Adj", "Adv", "Prep", "Scon", "Art"):
if lyr in sent and sent[lyr]:
for a, b, lbl in sent[lyr]:
if (
lbl == "X"
): # skip this, which covers an entire phrase typically containing the target and all its FEs
# (but do display the Gov)
continue
if any(1 for x, y, felbl in sent.FE[0] if x <= a < y or a <= x < b):
# overlap between one of the POS-specific layers and first FE layer
posspec_separate = (
True
) # show POS-specific layers on a separate line
posspec.append(
(a, b, lbl.lower().replace("-", ""))
) # lowercase Cop=>cop, Non-Asp=>nonasp, etc. to distinguish from FE names
if posspec_separate:
POSSPEC = _annotation_ascii_FE_layer(posspec, {}, feAbbrevs)
FE1 = _annotation_ascii_FE_layer(
sorted(sent.FE[0] + (posspec if not posspec_separate else [])),
sent.FE[1],
feAbbrevs,
)
FE2 = FE3 = None
if "FE2" in sent:
FE2 = _annotation_ascii_FE_layer(sent.FE2[0], sent.FE2[1], feAbbrevs)
if "FE3" in sent:
FE3 = _annotation_ascii_FE_layer(sent.FE3[0], sent.FE3[1], feAbbrevs)
for i, j in sent.Target:
FE1span, FE1name, FE1exp = FE1
if len(FE1span) < j:
FE1span += " " * (j - len(FE1span))
if len(FE1name) < j:
FE1name += " " * (j - len(FE1name))
FE1[1] = FE1name
FE1[0] = (
FE1span[:i] + FE1span[i:j].replace(" ", "*").replace("-", "=") + FE1span[j:]
)
long_lines = [sent.text]
if posspec_separate:
long_lines.extend(POSSPEC[:2])
long_lines.extend([FE1[0], FE1[1] + FE1[2]]) # lines with no length limit
if FE2:
long_lines.extend([FE2[0], FE2[1] + FE2[2]])
if FE3:
long_lines.extend([FE3[0], FE3[1] + FE3[2]])
long_lines.append("")
outstr = "\n".join(
map("\n".join, zip_longest(*mimic_wrap(long_lines), fillvalue=" "))
)
if feAbbrevs:
outstr += "(" + ", ".join("=".join(pair) for pair in feAbbrevs.items()) + ")"
assert len(feAbbrevs) == len(dict(feAbbrevs)), "Abbreviation clash"
outstr += "\n"
return outstr
def _pretty_fe(fe):
"""
Helper function for pretty-printing a frame element.
:param fe: The frame element to be printed.
:type fe: AttrDict
:return: A nicely formated string representation of the frame element.
:rtype: str
"""
fekeys = fe.keys()
outstr = ""
outstr += "frame element ({0.ID}): {0.name}\n of {1.name}({1.ID})\n".format(
fe, fe.frame
)
if "definition" in fekeys:
outstr += "[definition]\n"
outstr += _pretty_longstring(fe.definition, " ")
if "abbrev" in fekeys:
outstr += "[abbrev] {0}\n".format(fe.abbrev)
if "coreType" in fekeys:
outstr += "[coreType] {0}\n".format(fe.coreType)
if "requiresFE" in fekeys:
outstr += "[requiresFE] "
if fe.requiresFE is None:
outstr += "<None>\n"
else:
outstr += "{0}({1})\n".format(fe.requiresFE.name, fe.requiresFE.ID)
if "excludesFE" in fekeys:
outstr += "[excludesFE] "
if fe.excludesFE is None:
outstr += "<None>\n"
else:
outstr += "{0}({1})\n".format(fe.excludesFE.name, fe.excludesFE.ID)
if "semType" in fekeys:
outstr += "[semType] "
if fe.semType is None:
outstr += "<None>\n"
else:
outstr += "\n " + "{0}({1})".format(fe.semType.name, fe.semType.ID) + "\n"
return outstr
def _pretty_frame(frame):
"""
Helper function for pretty-printing a frame.
:param frame: The frame to be printed.
:type frame: AttrDict
:return: A nicely formated string representation of the frame.
:rtype: str
"""
outstr = ""
outstr += "frame ({0.ID}): {0.name}\n\n".format(frame)
outstr += "[URL] {0}\n\n".format(frame.URL)
outstr += "[definition]\n"
outstr += _pretty_longstring(frame.definition, " ") + "\n"
outstr += "[semTypes] {0} semantic types\n".format(len(frame.semTypes))
outstr += (
" " * (len(frame.semTypes) > 0)
+ ", ".join("{0}({1})".format(x.name, x.ID) for x in frame.semTypes)
+ "\n" * (len(frame.semTypes) > 0)
)
outstr += "\n[frameRelations] {0} frame relations\n".format(
len(frame.frameRelations)
)
outstr += " " + "\n ".join(repr(frel) for frel in frame.frameRelations) + "\n"
outstr += "\n[lexUnit] {0} lexical units\n".format(len(frame.lexUnit))
lustrs = []
for luName, lu in sorted(frame.lexUnit.items()):
tmpstr = "{0} ({1})".format(luName, lu.ID)
lustrs.append(tmpstr)
outstr += "{0}\n".format(_pretty_longstring(", ".join(lustrs), prefix=" "))
outstr += "\n[FE] {0} frame elements\n".format(len(frame.FE))
fes = {}
for feName, fe in sorted(frame.FE.items()):
try:
fes[fe.coreType].append("{0} ({1})".format(feName, fe.ID))
except KeyError:
fes[fe.coreType] = []
fes[fe.coreType].append("{0} ({1})".format(feName, fe.ID))
for ct in sorted(
fes.keys(),
key=lambda ct2: [
"Core",
"Core-Unexpressed",
"Peripheral",
"Extra-Thematic",
].index(ct2),
):
outstr += "{0:>16}: {1}\n".format(ct, ", ".join(sorted(fes[ct])))
outstr += "\n[FEcoreSets] {0} frame element core sets\n".format(
len(frame.FEcoreSets)
)
outstr += (
" "
+ "\n ".join(
", ".join([x.name for x in coreSet]) for coreSet in frame.FEcoreSets
)
+ "\n"
)
return outstr
class FramenetError(Exception):
"""An exception class for framenet-related errors."""
class AttrDict(dict):
"""A class that wraps a dict and allows accessing the keys of the
dict as if they were attributes. Taken from here:
http://stackoverflow.com/a/14620633/8879
>>> foo = {'a':1, 'b':2, 'c':3}
>>> bar = AttrDict(foo)
>>> pprint(dict(bar))
{'a': 1, 'b': 2, 'c': 3}
>>> bar.b
2
>>> bar.d = 4
>>> pprint(dict(bar))
{'a': 1, 'b': 2, 'c': 3, 'd': 4}
"""
def __init__(self, *args, **kwargs):
super(AttrDict, self).__init__(*args, **kwargs)
# self.__dict__ = self
def __setattr__(self, name, value):
self[name] = value
def __getattr__(self, name):
if name == "_short_repr":
return self._short_repr
return self[name]
def __getitem__(self, name):
v = super(AttrDict, self).__getitem__(name)
if isinstance(v, Future):
return v._data()
return v
def _short_repr(self):
if "_type" in self:
if self["_type"].endswith("relation"):
return self.__repr__()
try:
return "<{0} ID={1} name={2}>".format(
self["_type"], self["ID"], self["name"]
)
except KeyError:
try: # no ID--e.g., for _type=lusubcorpus
return "<{0} name={1}>".format(self["_type"], self["name"])
except KeyError: # no name--e.g., for _type=lusentence
return "<{0} ID={1}>".format(self["_type"], self["ID"])
else:
return self.__repr__()
def _str(self):
outstr = ""
if "_type" not in self:
outstr = _pretty_any(self)
elif self["_type"] == "frame":
outstr = _pretty_frame(self)
elif self["_type"] == "fe":
outstr = _pretty_fe(self)
elif self["_type"] == "lu":
outstr = _pretty_lu(self)
elif self["_type"] == "luexemplars": # list of ALL exemplars for LU
outstr = _pretty_exemplars(self, self[0].LU)
elif (
self["_type"] == "fulltext_annotation"
): # list of all sentences for full-text doc
outstr = _pretty_fulltext_sentences(self)
elif self["_type"] == "lusentence":
outstr = _pretty_annotation(self)
elif self["_type"] == "fulltext_sentence":
outstr = _pretty_fulltext_sentence(self)
elif self["_type"] in ("luannotationset", "fulltext_annotationset"):
outstr = _pretty_annotation(self, aset_level=True)
elif self["_type"] == "posannotationset":
outstr = _pretty_pos(self)
elif self["_type"] == "semtype":
outstr = _pretty_semtype(self)
elif self["_type"] == "framerelationtype":
outstr = _pretty_frame_relation_type(self)
elif self["_type"] == "framerelation":
outstr = _pretty_frame_relation(self)
elif self["_type"] == "ferelation":
outstr = _pretty_fe_relation(self)
else:
outstr = _pretty_any(self)
# ensure result is unicode string prior to applying the
# decorator (because non-ASCII characters
# could in principle occur in the data and would trigger an encoding error when
# passed as arguments to str.format()).
# assert isinstance(outstr, unicode) # not in Python 3.2
return outstr
def __str__(self):
return self._str()
def __repr__(self):
return self.__str__()
class SpecialList(list):
"""
A list subclass which adds a '_type' attribute for special printing
(similar to an AttrDict, though this is NOT an AttrDict subclass).
"""
def __init__(self, typ, *args, **kwargs):
super(SpecialList, self).__init__(*args, **kwargs)
self._type = typ
def _str(self):
outstr = ""
assert self._type
if len(self) == 0:
outstr = "[]"
elif self._type == "luexemplars": # list of ALL exemplars for LU
outstr = _pretty_exemplars(self, self[0].LU)
else:
assert False, self._type
return outstr
def __str__(self):
return self._str()
def __repr__(self):
return self.__str__()
class Future(object):
"""
Wraps and acts as a proxy for a value to be loaded lazily (on demand).
Adapted from https://gist.github.com/sergey-miryanov/2935416
"""
def __init__(self, loader, *args, **kwargs):
"""
:param loader: when called with no arguments, returns the value to be stored
:type loader: callable
"""
super(Future, self).__init__(*args, **kwargs)
self._loader = loader
self._d = None
def _data(self):
if callable(self._loader):
self._d = self._loader()
self._loader = None # the data is now cached
return self._d
def __nonzero__(self):
return bool(self._data())
def __len__(self):
return len(self._data())
def __setitem__(self, key, value):
return self._data().__setitem__(key, value)
def __getitem__(self, key):
return self._data().__getitem__(key)
def __getattr__(self, key):
return self._data().__getattr__(key)
def __str__(self):
return self._data().__str__()
def __repr__(self):
return self._data().__repr__()
class PrettyDict(AttrDict):
"""
Displays an abbreviated repr of values where possible.
Inherits from AttrDict, so a callable value will
be lazily converted to an actual value.
"""
def __init__(self, *args, **kwargs):
_BREAK_LINES = kwargs.pop("breakLines", False)
super(PrettyDict, self).__init__(*args, **kwargs)
dict.__setattr__(self, "_BREAK_LINES", _BREAK_LINES)
def __repr__(self):
parts = []
for k, v in sorted(self.items()):
kv = repr(k) + ": "
try:
kv += v._short_repr()
except AttributeError:
kv += repr(v)
parts.append(kv)
return "{" + (",\n " if self._BREAK_LINES else ", ").join(parts) + "}"
class PrettyList(list):
"""
Displays an abbreviated repr of only the first several elements, not the whole list.
"""
# from nltk.util
def __init__(self, *args, **kwargs):
self._MAX_REPR_SIZE = kwargs.pop("maxReprSize", 60)
self._BREAK_LINES = kwargs.pop("breakLines", False)
super(PrettyList, self).__init__(*args, **kwargs)
def __repr__(self):
"""
Return a string representation for this corpus view that is
similar to a list's representation; but if it would be more
than 60 characters long, it is truncated.
"""
pieces = []
length = 5
for elt in self:
pieces.append(
elt._short_repr()
) # key difference from inherited version: call to _short_repr()
length += len(pieces[-1]) + 2
if self._MAX_REPR_SIZE and length > self._MAX_REPR_SIZE and len(pieces) > 2:
return "[%s, ...]" % str(
",\n " if self._BREAK_LINES else ", "
).join(pieces[:-1])
return "[%s]" % str(",\n " if self._BREAK_LINES else ", ").join(pieces)
class PrettyLazyMap(LazyMap):
"""
Displays an abbreviated repr of only the first several elements, not the whole list.
"""
# from nltk.util
_MAX_REPR_SIZE = 60
def __repr__(self):
"""
Return a string representation for this corpus view that is
similar to a list's representation; but if it would be more
than 60 characters long, it is truncated.
"""
pieces = []
length = 5
for elt in self:
pieces.append(
elt._short_repr()
) # key difference from inherited version: call to _short_repr()
length += len(pieces[-1]) + 2
if length > self._MAX_REPR_SIZE and len(pieces) > 2:
return "[%s, ...]" % str(", ").join(pieces[:-1])
return "[%s]" % str(", ").join(pieces)
class PrettyLazyIteratorList(LazyIteratorList):
"""
Displays an abbreviated repr of only the first several elements, not the whole list.
"""
# from nltk.util
_MAX_REPR_SIZE = 60
def __repr__(self):
"""
Return a string representation for this corpus view that is
similar to a list's representation; but if it would be more
than 60 characters long, it is truncated.
"""
pieces = []
length = 5
for elt in self:
pieces.append(
elt._short_repr()
) # key difference from inherited version: call to _short_repr()
length += len(pieces[-1]) + 2
if length > self._MAX_REPR_SIZE and len(pieces) > 2:
return "[%s, ...]" % str(", ").join(pieces[:-1])
return "[%s]" % str(", ").join(pieces)
class PrettyLazyConcatenation(LazyConcatenation):
"""
Displays an abbreviated repr of only the first several elements, not the whole list.
"""
# from nltk.util
_MAX_REPR_SIZE = 60
def __repr__(self):
"""
Return a string representation for this corpus view that is
similar to a list's representation; but if it would be more
than 60 characters long, it is truncated.
"""
pieces = []
length = 5
for elt in self:
pieces.append(
elt._short_repr()
) # key difference from inherited version: call to _short_repr()
length += len(pieces[-1]) + 2
if length > self._MAX_REPR_SIZE and len(pieces) > 2:
return "[%s, ...]" % str(", ").join(pieces[:-1])
return "[%s]" % str(", ").join(pieces)
def __add__(self, other):
"""Return a list concatenating self with other."""
return PrettyLazyIteratorList(itertools.chain(self, other))
def __radd__(self, other):
"""Return a list concatenating other with self."""
return PrettyLazyIteratorList(itertools.chain(other, self))
class FramenetCorpusReader(XMLCorpusReader):
"""A corpus reader for the Framenet Corpus.
>>> from nltk.corpus import framenet as fn
>>> fn.lu(3238).frame.lexUnit['glint.v'] is fn.lu(3238)
True
>>> fn.frame_by_name('Replacing') is fn.lus('replace.v')[0].frame
True
>>> fn.lus('prejudice.n')[0].frame.frameRelations == fn.frame_relations('Partiality')
True
"""
_bad_statuses = ["Problem"]
"""
When loading LUs for a frame, those whose status is in this list will be ignored.
Due to caching, if user code modifies this, it should do so before loading any data.
'Problem' should always be listed for FrameNet 1.5, as these LUs are not included
in the XML index.
"""
_warnings = False
def warnings(self, v):
"""Enable or disable warnings of data integrity issues as they are encountered.
If v is truthy, warnings will be enabled.
(This is a function rather than just an attribute/property to ensure that if
enabling warnings is the first action taken, the corpus reader is instantiated first.)
"""
self._warnings = v
def __init__(self, root, fileids):
XMLCorpusReader.__init__(self, root, fileids)
# framenet corpus sub dirs
# sub dir containing the xml files for frames
self._frame_dir = "frame"
# sub dir containing the xml files for lexical units
self._lu_dir = "lu"
# sub dir containing the xml files for fulltext annotation files
self._fulltext_dir = "fulltext"
# location of latest development version of FrameNet
self._fnweb_url = "https://framenet2.icsi.berkeley.edu/fnReports/data"
# Indexes used for faster look-ups
self._frame_idx = None
self._cached_frames = {} # name -> ID
self._lu_idx = None
self._fulltext_idx = None
self._semtypes = None
self._freltyp_idx = None # frame relation types (Inheritance, Using, etc.)
self._frel_idx = None # frame-to-frame relation instances
self._ferel_idx = None # FE-to-FE relation instances
self._frel_f_idx = None # frame-to-frame relations associated with each frame
def help(self, attrname=None):
"""Display help information summarizing the main methods."""
if attrname is not None:
return help(self.__getattribute__(attrname))
# No need to mention frame_by_name() or frame_by_id(),
# as it's easier to just call frame().
# Also not mentioning lu_basic().
msg = """
Citation: Nathan Schneider and Chuck Wooters (2017),
"The NLTK FrameNet API: Designing for Discoverability with a Rich Linguistic Resource".
Proceedings of EMNLP: System Demonstrations. https://arxiv.org/abs/1703.07438
Use the following methods to access data in FrameNet.
Provide a method name to `help()` for more information.
FRAMES
======
frame() to look up a frame by its exact name or ID
frames() to get frames matching a name pattern
frames_by_lemma() to get frames containing an LU matching a name pattern
frame_ids_and_names() to get a mapping from frame IDs to names
FRAME ELEMENTS
==============
fes() to get frame elements (a.k.a. roles) matching a name pattern, optionally constrained
by a frame name pattern
LEXICAL UNITS
=============
lu() to look up an LU by its ID
lus() to get lexical units matching a name pattern, optionally constrained by frame
lu_ids_and_names() to get a mapping from LU IDs to names
RELATIONS
=========
frame_relation_types() to get the different kinds of frame-to-frame relations
(Inheritance, Subframe, Using, etc.).
frame_relations() to get the relation instances, optionally constrained by
frame(s) or relation type
fe_relations() to get the frame element pairs belonging to a frame-to-frame relation
SEMANTIC TYPES
==============
semtypes() to get the different kinds of semantic types that can be applied to
FEs, LUs, and entire frames
semtype() to look up a particular semtype by name, ID, or abbreviation
semtype_inherits() to check whether two semantic types have a subtype-supertype
relationship in the semtype hierarchy
propagate_semtypes() to apply inference rules that distribute semtypes over relations
between FEs
ANNOTATIONS
===========
annotations() to get annotation sets, in which a token in a sentence is annotated
with a lexical unit in a frame, along with its frame elements and their syntactic properties;
can be constrained by LU name pattern and limited to lexicographic exemplars or full-text.
Sentences of full-text annotation can have multiple annotation sets.
sents() to get annotated sentences illustrating one or more lexical units
exemplars() to get sentences of lexicographic annotation, most of which have
just 1 annotation set; can be constrained by LU name pattern, frame, and overt FE(s)
doc() to look up a document of full-text annotation by its ID
docs() to get documents of full-text annotation that match a name pattern
docs_metadata() to get metadata about all full-text documents without loading them
ft_sents() to iterate over sentences of full-text annotation
UTILITIES
=========
buildindexes() loads metadata about all frames, LUs, etc. into memory to avoid
delay when one is accessed for the first time. It does not load annotations.
readme() gives the text of the FrameNet README file
warnings(True) to display corpus consistency warnings when loading data
"""
print(msg)
def _buildframeindex(self):
# The total number of Frames in Framenet is fairly small (~1200) so
# this index should not be very large
if not self._frel_idx:
self._buildrelationindex() # always load frame relations before frames,
# otherwise weird ordering effects might result in incomplete information
self._frame_idx = {}
for f in XMLCorpusView(
self.abspath("frameIndex.xml"), "frameIndex/frame", self._handle_elt
):
self._frame_idx[f["ID"]] = f
def _buildcorpusindex(self):
# The total number of fulltext annotated documents in Framenet
# is fairly small (~90) so this index should not be very large
self._fulltext_idx = {}
for doclist in XMLCorpusView(
self.abspath("fulltextIndex.xml"),
"fulltextIndex/corpus",
self._handle_fulltextindex_elt,
):
for doc in doclist:
self._fulltext_idx[doc.ID] = doc
def _buildluindex(self):
# The number of LUs in Framenet is about 13,000 so this index
# should not be very large
self._lu_idx = {}
for lu in XMLCorpusView(
self.abspath("luIndex.xml"), "luIndex/lu", self._handle_elt
):
self._lu_idx[
lu["ID"]
] = lu # populate with LU index entries. if any of these
# are looked up they will be replaced by full LU objects.
def _buildrelationindex(self):
# print('building relation index...', file=sys.stderr)
freltypes = PrettyList(
x
for x in XMLCorpusView(
self.abspath("frRelation.xml"),
"frameRelations/frameRelationType",
self._handle_framerelationtype_elt,
)
)
self._freltyp_idx = {}
self._frel_idx = {}
self._frel_f_idx = defaultdict(set)
self._ferel_idx = {}
for freltyp in freltypes:
self._freltyp_idx[freltyp.ID] = freltyp
for frel in freltyp.frameRelations:
supF = frel.superFrame = frel[freltyp.superFrameName] = Future(
(lambda fID: lambda: self.frame_by_id(fID))(frel.supID)
)
subF = frel.subFrame = frel[freltyp.subFrameName] = Future(
(lambda fID: lambda: self.frame_by_id(fID))(frel.subID)
)
self._frel_idx[frel.ID] = frel
self._frel_f_idx[frel.supID].add(frel.ID)
self._frel_f_idx[frel.subID].add(frel.ID)
for ferel in frel.feRelations:
ferel.superFrame = supF
ferel.subFrame = subF
ferel.superFE = Future(
(lambda fer: lambda: fer.superFrame.FE[fer.superFEName])(ferel)
)
ferel.subFE = Future(
(lambda fer: lambda: fer.subFrame.FE[fer.subFEName])(ferel)
)
self._ferel_idx[ferel.ID] = ferel
# print('...done building relation index', file=sys.stderr)
def _warn(self, *message, **kwargs):
if self._warnings:
kwargs.setdefault("file", sys.stderr)
print(*message, **kwargs)
def readme(self):
"""
Return the contents of the corpus README.txt (or README) file.
"""
try:
return self.open("README.txt").read()
except IOError:
return self.open("README").read()
def buildindexes(self):
"""
Build the internal indexes to make look-ups faster.
"""
# Frames
self._buildframeindex()
# LUs
self._buildluindex()
# Fulltext annotation corpora index
self._buildcorpusindex()
# frame and FE relations
self._buildrelationindex()
def doc(self, fn_docid):
"""
Returns the annotated document whose id number is
``fn_docid``. This id number can be obtained by calling the
Documents() function.
The dict that is returned from this function will contain the
following keys:
- '_type' : 'fulltextannotation'
- 'sentence' : a list of sentences in the document
- Each item in the list is a dict containing the following keys:
- 'ID' : the ID number of the sentence
- '_type' : 'sentence'
- 'text' : the text of the sentence
- 'paragNo' : the paragraph number
- 'sentNo' : the sentence number
- 'docID' : the document ID number
- 'corpID' : the corpus ID number
- 'aPos' : the annotation position
- 'annotationSet' : a list of annotation layers for the sentence
- Each item in the list is a dict containing the following keys:
- 'ID' : the ID number of the annotation set
- '_type' : 'annotationset'
- 'status' : either 'MANUAL' or 'UNANN'
- 'luName' : (only if status is 'MANUAL')
- 'luID' : (only if status is 'MANUAL')
- 'frameID' : (only if status is 'MANUAL')
- 'frameName': (only if status is 'MANUAL')
- 'layer' : a list of labels for the layer
- Each item in the layer is a dict containing the
following keys:
- '_type': 'layer'
- 'rank'
- 'name'
- 'label' : a list of labels in the layer
- Each item is a dict containing the following keys:
- 'start'
- 'end'
- 'name'
- 'feID' (optional)
:param fn_docid: The Framenet id number of the document
:type fn_docid: int
:return: Information about the annotated document
:rtype: dict
"""
try:
xmlfname = self._fulltext_idx[fn_docid].filename
except TypeError: # happens when self._fulltext_idx == None
# build the index
self._buildcorpusindex()
xmlfname = self._fulltext_idx[fn_docid].filename
except KeyError: # probably means that fn_docid was not in the index
raise FramenetError("Unknown document id: {0}".format(fn_docid))
# construct the path name for the xml file containing the document info
locpath = os.path.join("{0}".format(self._root), self._fulltext_dir, xmlfname)
# Grab the top-level xml element containing the fulltext annotation
elt = XMLCorpusView(locpath, "fullTextAnnotation")[0]
info = self._handle_fulltextannotation_elt(elt)
# add metadata
for k, v in self._fulltext_idx[fn_docid].items():
info[k] = v
return info
def frame_by_id(self, fn_fid, ignorekeys=[]):
"""
Get the details for the specified Frame using the frame's id
number.
Usage examples:
>>> from nltk.corpus import framenet as fn
>>> f = fn.frame_by_id(256)
>>> f.ID
256
>>> f.name
'Medical_specialties'
>>> f.definition
"This frame includes words that name ..."
:param fn_fid: The Framenet id number of the frame
:type fn_fid: int
:param ignorekeys: The keys to ignore. These keys will not be
included in the output. (optional)
:type ignorekeys: list(str)
:return: Information about a frame
:rtype: dict
Also see the ``frame()`` function for details about what is
contained in the dict that is returned.
"""
# get the name of the frame with this id number
try:
fentry = self._frame_idx[fn_fid]
if "_type" in fentry:
return fentry # full frame object is cached
name = fentry["name"]
except TypeError:
self._buildframeindex()
name = self._frame_idx[fn_fid]["name"]
except KeyError:
raise FramenetError("Unknown frame id: {0}".format(fn_fid))
return self.frame_by_name(name, ignorekeys, check_cache=False)
def frame_by_name(self, fn_fname, ignorekeys=[], check_cache=True):
"""
Get the details for the specified Frame using the frame's name.
Usage examples:
>>> from nltk.corpus import framenet as fn
>>> f = fn.frame_by_name('Medical_specialties')
>>> f.ID
256
>>> f.name
'Medical_specialties'
>>> f.definition
"This frame includes words that name ..."
:param fn_fname: The name of the frame
:type fn_fname: str
:param ignorekeys: The keys to ignore. These keys will not be
included in the output. (optional)
:type ignorekeys: list(str)
:return: Information about a frame
:rtype: dict
Also see the ``frame()`` function for details about what is
contained in the dict that is returned.
"""
if check_cache and fn_fname in self._cached_frames:
return self._frame_idx[self._cached_frames[fn_fname]]
elif not self._frame_idx:
self._buildframeindex()
# construct the path name for the xml file containing the Frame info
locpath = os.path.join(
"{0}".format(self._root), self._frame_dir, fn_fname + ".xml"
)
# print(locpath, file=sys.stderr)
# Grab the xml for the frame
try:
elt = XMLCorpusView(locpath, "frame")[0]
except IOError:
raise FramenetError("Unknown frame: {0}".format(fn_fname))
fentry = self._handle_frame_elt(elt, ignorekeys)
assert fentry
fentry.URL = self._fnweb_url + "/" + self._frame_dir + "/" + fn_fname + ".xml"
# INFERENCE RULE: propagate lexical semtypes from the frame to all its LUs
for st in fentry.semTypes:
if st.rootType.name == "Lexical_type":
for lu in fentry.lexUnit.values():
if not any(
x is st for x in lu.semTypes
): # identity containment check
lu.semTypes.append(st)
self._frame_idx[fentry.ID] = fentry
self._cached_frames[fentry.name] = fentry.ID
"""
# now set up callables to resolve the LU pointers lazily.
# (could also do this here--caching avoids infinite recursion.)
for luName,luinfo in fentry.lexUnit.items():
fentry.lexUnit[luName] = (lambda luID: Future(lambda: self.lu(luID)))(luinfo.ID)
"""
return fentry
def frame(self, fn_fid_or_fname, ignorekeys=[]):
"""
Get the details for the specified Frame using the frame's name
or id number.
Usage examples:
>>> from nltk.corpus import framenet as fn
>>> f = fn.frame(256)
>>> f.name
'Medical_specialties'
>>> f = fn.frame('Medical_specialties')
>>> f.ID
256
>>> # ensure non-ASCII character in definition doesn't trigger an encoding error:
>>> fn.frame('Imposing_obligation')
frame (1494): Imposing_obligation...
The dict that is returned from this function will contain the
following information about the Frame:
- 'name' : the name of the Frame (e.g. 'Birth', 'Apply_heat', etc.)
- 'definition' : textual definition of the Frame
- 'ID' : the internal ID number of the Frame
- 'semTypes' : a list of semantic types for this frame
- Each item in the list is a dict containing the following keys:
- 'name' : can be used with the semtype() function
- 'ID' : can be used with the semtype() function
- 'lexUnit' : a dict containing all of the LUs for this frame.
The keys in this dict are the names of the LUs and
the value for each key is itself a dict containing
info about the LU (see the lu() function for more info.)
- 'FE' : a dict containing the Frame Elements that are part of this frame
The keys in this dict are the names of the FEs (e.g. 'Body_system')
and the values are dicts containing the following keys
- 'definition' : The definition of the FE
- 'name' : The name of the FE e.g. 'Body_system'
- 'ID' : The id number
- '_type' : 'fe'
- 'abbrev' : Abbreviation e.g. 'bod'
- 'coreType' : one of "Core", "Peripheral", or "Extra-Thematic"
- 'semType' : if not None, a dict with the following two keys:
- 'name' : name of the semantic type. can be used with
the semtype() function
- 'ID' : id number of the semantic type. can be used with
the semtype() function
- 'requiresFE' : if not None, a dict with the following two keys:
- 'name' : the name of another FE in this frame
- 'ID' : the id of the other FE in this frame
- 'excludesFE' : if not None, a dict with the following two keys:
- 'name' : the name of another FE in this frame
- 'ID' : the id of the other FE in this frame
- 'frameRelation' : a list of objects describing frame relations
- 'FEcoreSets' : a list of Frame Element core sets for this frame
- Each item in the list is a list of FE objects
:param fn_fid_or_fname: The Framenet name or id number of the frame
:type fn_fid_or_fname: int or str
:param ignorekeys: The keys to ignore. These keys will not be
included in the output. (optional)
:type ignorekeys: list(str)
:return: Information about a frame
:rtype: dict
"""
# get the frame info by name or id number
if isinstance(fn_fid_or_fname, str):
f = self.frame_by_name(fn_fid_or_fname, ignorekeys)
else:
f = self.frame_by_id(fn_fid_or_fname, ignorekeys)
return f
def frames_by_lemma(self, pat):
"""
Returns a list of all frames that contain LUs in which the
``name`` attribute of the LU matchs the given regular expression
``pat``. Note that LU names are composed of "lemma.POS", where
the "lemma" part can be made up of either a single lexeme
(e.g. 'run') or multiple lexemes (e.g. 'a little').
Note: if you are going to be doing a lot of this type of
searching, you'd want to build an index that maps from lemmas to
frames because each time frames_by_lemma() is called, it has to
search through ALL of the frame XML files in the db.
>>> from nltk.corpus import framenet as fn
>>> from nltk.corpus.reader.framenet import PrettyList
>>> PrettyList(sorted(fn.frames_by_lemma(r'(?i)a little'), key=itemgetter('ID'))) # doctest: +ELLIPSIS
[<frame ID=189 name=Quanti...>, <frame ID=2001 name=Degree>]
:return: A list of frame objects.
:rtype: list(AttrDict)
"""
return PrettyList(
f
for f in self.frames()
if any(re.search(pat, luName) for luName in f.lexUnit)
)
def lu_basic(self, fn_luid):
"""
Returns basic information about the LU whose id is
``fn_luid``. This is basically just a wrapper around the
``lu()`` function with "subCorpus" info excluded.
>>> from nltk.corpus import framenet as fn
>>> lu = PrettyDict(fn.lu_basic(256), breakLines=True)
>>> # ellipses account for differences between FN 1.5 and 1.7
>>> lu # doctest: +ELLIPSIS
{'ID': 256,
'POS': 'V',
'URL': 'https://framenet2.icsi.berkeley.edu/fnReports/data/lu/lu256.xml',
'_type': 'lu',
'cBy': ...,
'cDate': '02/08/2001 01:27:50 PST Thu',
'definition': 'COD: be aware of beforehand; predict.',
'definitionMarkup': 'COD: be aware of beforehand; predict.',
'frame': <frame ID=26 name=Expectation>,
'lemmaID': 15082,
'lexemes': [{'POS': 'V', 'breakBefore': 'false', 'headword': 'false', 'name': 'foresee', 'order': 1}],
'name': 'foresee.v',
'semTypes': [],
'sentenceCount': {'annotated': ..., 'total': ...},
'status': 'FN1_Sent'}
:param fn_luid: The id number of the desired LU
:type fn_luid: int
:return: Basic information about the lexical unit
:rtype: dict
"""
return self.lu(fn_luid, ignorekeys=["subCorpus", "exemplars"])
def lu(self, fn_luid, ignorekeys=[], luName=None, frameID=None, frameName=None):
"""
Access a lexical unit by its ID. luName, frameID, and frameName are used
only in the event that the LU does not have a file in the database
(which is the case for LUs with "Problem" status); in this case,
a placeholder LU is created which just contains its name, ID, and frame.
Usage examples:
>>> from nltk.corpus import framenet as fn
>>> fn.lu(256).name
'foresee.v'
>>> fn.lu(256).definition
'COD: be aware of beforehand; predict.'
>>> fn.lu(256).frame.name
'Expectation'
>>> pprint(list(map(PrettyDict, fn.lu(256).lexemes)))
[{'POS': 'V', 'breakBefore': 'false', 'headword': 'false', 'name': 'foresee', 'order': 1}]
>>> fn.lu(227).exemplars[23]
exemplar sentence (352962):
[sentNo] 0
[aPos] 59699508
<BLANKLINE>
[LU] (227) guess.v in Coming_to_believe
<BLANKLINE>
[frame] (23) Coming_to_believe
<BLANKLINE>
[annotationSet] 2 annotation sets
<BLANKLINE>
[POS] 18 tags
<BLANKLINE>
[POS_tagset] BNC
<BLANKLINE>
[GF] 3 relations
<BLANKLINE>
[PT] 3 phrases
<BLANKLINE>
[Other] 1 entry
<BLANKLINE>
[text] + [Target] + [FE]
<BLANKLINE>
When he was inside the house , Culley noticed the characteristic
------------------
Content
<BLANKLINE>
he would n't have guessed at .
-- ******* --
Co C1 [Evidence:INI]
(Co=Cognizer, C1=Content)
<BLANKLINE>
<BLANKLINE>
The dict that is returned from this function will contain most of the
following information about the LU. Note that some LUs do not contain
all of these pieces of information - particularly 'totalAnnotated' and
'incorporatedFE' may be missing in some LUs:
- 'name' : the name of the LU (e.g. 'merger.n')
- 'definition' : textual definition of the LU
- 'ID' : the internal ID number of the LU
- '_type' : 'lu'
- 'status' : e.g. 'Created'
- 'frame' : Frame that this LU belongs to
- 'POS' : the part of speech of this LU (e.g. 'N')
- 'totalAnnotated' : total number of examples annotated with this LU
- 'incorporatedFE' : FE that incorporates this LU (e.g. 'Ailment')
- 'sentenceCount' : a dict with the following two keys:
- 'annotated': number of sentences annotated with this LU
- 'total' : total number of sentences with this LU
- 'lexemes' : a list of dicts describing the lemma of this LU.
Each dict in the list contains these keys:
- 'POS' : part of speech e.g. 'N'
- 'name' : either single-lexeme e.g. 'merger' or
multi-lexeme e.g. 'a little'
- 'order': the order of the lexeme in the lemma (starting from 1)
- 'headword': a boolean ('true' or 'false')
- 'breakBefore': Can this lexeme be separated from the previous lexeme?
Consider: "take over.v" as in:
Germany took over the Netherlands in 2 days.
Germany took the Netherlands over in 2 days.
In this case, 'breakBefore' would be "true" for the lexeme
"over". Contrast this with "take after.v" as in:
Mary takes after her grandmother.
*Mary takes her grandmother after.
In this case, 'breakBefore' would be "false" for the lexeme "after"
- 'lemmaID' : Can be used to connect lemmas in different LUs
- 'semTypes' : a list of semantic type objects for this LU
- 'subCorpus' : a list of subcorpora
- Each item in the list is a dict containing the following keys:
- 'name' :
- 'sentence' : a list of sentences in the subcorpus
- each item in the list is a dict with the following keys:
- 'ID':
- 'sentNo':
- 'text': the text of the sentence
- 'aPos':
- 'annotationSet': a list of annotation sets
- each item in the list is a dict with the following keys:
- 'ID':
- 'status':
- 'layer': a list of layers
- each layer is a dict containing the following keys:
- 'name': layer name (e.g. 'BNC')
- 'rank':
- 'label': a list of labels for the layer
- each label is a dict containing the following keys:
- 'start': start pos of label in sentence 'text' (0-based)
- 'end': end pos of label in sentence 'text' (0-based)
- 'name': name of label (e.g. 'NN1')
Under the hood, this implementation looks up the lexical unit information
in the *frame* definition file. That file does not contain
corpus annotations, so the LU files will be accessed on demand if those are
needed. In principle, valence patterns could be loaded here too,
though these are not currently supported.
:param fn_luid: The id number of the lexical unit
:type fn_luid: int
:param ignorekeys: The keys to ignore. These keys will not be
included in the output. (optional)
:type ignorekeys: list(str)
:return: All information about the lexical unit
:rtype: dict
"""
# look for this LU in cache
if not self._lu_idx:
self._buildluindex()
OOV = object()
luinfo = self._lu_idx.get(fn_luid, OOV)
if luinfo is OOV:
# LU not in the index. We create a placeholder by falling back to
# luName, frameID, and frameName. However, this will not be listed
# among the LUs for its frame.
self._warn(
"LU ID not found: {0} ({1}) in {2} ({3})".format(
luName, fn_luid, frameName, frameID
)
)
luinfo = AttrDict(
{
"_type": "lu",
"ID": fn_luid,
"name": luName,
"frameID": frameID,
"status": "Problem",
}
)
f = self.frame_by_id(luinfo.frameID)
assert f.name == frameName, (f.name, frameName)
luinfo["frame"] = f
self._lu_idx[fn_luid] = luinfo
elif "_type" not in luinfo:
# we only have an index entry for the LU. loading the frame will replace this.
f = self.frame_by_id(luinfo.frameID)
luinfo = self._lu_idx[fn_luid]
if ignorekeys:
return AttrDict(
dict((k, v) for k, v in luinfo.items() if k not in ignorekeys)
)
return luinfo
def _lu_file(self, lu, ignorekeys=[]):
"""
Augment the LU information that was loaded from the frame file
with additional information from the LU file.
"""
fn_luid = lu.ID
fname = "lu{0}.xml".format(fn_luid)
locpath = os.path.join("{0}".format(self._root), self._lu_dir, fname)
# print(locpath, file=sys.stderr)
if not self._lu_idx:
self._buildluindex()
try:
elt = XMLCorpusView(locpath, "lexUnit")[0]
except IOError:
raise FramenetError("Unknown LU id: {0}".format(fn_luid))
lu2 = self._handle_lexunit_elt(elt, ignorekeys)
lu.URL = self._fnweb_url + "/" + self._lu_dir + "/" + fname
lu.subCorpus = lu2.subCorpus
lu.exemplars = SpecialList(
"luexemplars", [sent for subc in lu.subCorpus for sent in subc.sentence]
)
for sent in lu.exemplars:
sent["LU"] = lu
sent["frame"] = lu.frame
for aset in sent.annotationSet:
aset["LU"] = lu
aset["frame"] = lu.frame
return lu
def _loadsemtypes(self):
"""Create the semantic types index."""
self._semtypes = AttrDict()
semtypeXML = [
x
for x in XMLCorpusView(
self.abspath("semTypes.xml"),
"semTypes/semType",
self._handle_semtype_elt,
)
]
for st in semtypeXML:
n = st["name"]
a = st["abbrev"]
i = st["ID"]
# Both name and abbrev should be able to retrieve the
# ID. The ID will retrieve the semantic type dict itself.
self._semtypes[n] = i
self._semtypes[a] = i
self._semtypes[i] = st
# now that all individual semtype XML is loaded, we can link them together
roots = []
for st in self.semtypes():
if st.superType:
st.superType = self.semtype(st.superType.supID)
st.superType.subTypes.append(st)
else:
if st not in roots:
roots.append(st)
st.rootType = st
queue = list(roots)
assert queue
while queue:
st = queue.pop(0)
for child in st.subTypes:
child.rootType = st.rootType
queue.append(child)
# self.propagate_semtypes() # apply inferencing over FE relations
def propagate_semtypes(self):
"""
Apply inference rules to distribute semtypes over relations between FEs.
For FrameNet 1.5, this results in 1011 semtypes being propagated.
(Not done by default because it requires loading all frame files,
which takes several seconds. If this needed to be fast, it could be rewritten
to traverse the neighboring relations on demand for each FE semtype.)
>>> from nltk.corpus import framenet as fn
>>> x = sum(1 for f in fn.frames() for fe in f.FE.values() if fe.semType)
>>> fn.propagate_semtypes()
>>> y = sum(1 for f in fn.frames() for fe in f.FE.values() if fe.semType)
>>> y-x > 1000
True
"""
if not self._semtypes:
self._loadsemtypes()
if not self._ferel_idx:
self._buildrelationindex()
changed = True
i = 0
nPropagations = 0
while changed:
# make a pass and see if anything needs to be propagated
i += 1
changed = False
for ferel in self.fe_relations():
superST = ferel.superFE.semType
subST = ferel.subFE.semType
try:
if superST and superST is not subST:
# propagate downward
assert subST is None or self.semtype_inherits(subST, superST), (
superST.name,
ferel,
subST.name,
)
if subST is None:
ferel.subFE.semType = subST = superST
changed = True
nPropagations += 1
if (
ferel.type.name in ["Perspective_on", "Subframe", "Precedes"]
and subST
and subST is not superST
):
# propagate upward
assert superST is None, (superST.name, ferel, subST.name)
ferel.superFE.semType = superST = subST
changed = True
nPropagations += 1
except AssertionError as ex:
# bug in the data! ignore
# print(ex, file=sys.stderr)
continue
# print(i, nPropagations, file=sys.stderr)
def semtype(self, key):
"""
>>> from nltk.corpus import framenet as fn
>>> fn.semtype(233).name
'Temperature'
>>> fn.semtype(233).abbrev
'Temp'
>>> fn.semtype('Temperature').ID
233
:param key: The name, abbreviation, or id number of the semantic type
:type key: string or int
:return: Information about a semantic type
:rtype: dict
"""
if isinstance(key, int):
stid = key
else:
try:
stid = self._semtypes[key]
except TypeError:
self._loadsemtypes()
stid = self._semtypes[key]
try:
st = self._semtypes[stid]
except TypeError:
self._loadsemtypes()
st = self._semtypes[stid]
return st
def semtype_inherits(self, st, superST):
if not isinstance(st, dict):
st = self.semtype(st)
if not isinstance(superST, dict):
superST = self.semtype(superST)
par = st.superType
while par:
if par is superST:
return True
par = par.superType
return False
def frames(self, name=None):
"""
Obtain details for a specific frame.
>>> from nltk.corpus import framenet as fn
>>> len(fn.frames()) in (1019, 1221) # FN 1.5 and 1.7, resp.
True
>>> x = PrettyList(fn.frames(r'(?i)crim'), maxReprSize=0, breakLines=True)
>>> x.sort(key=itemgetter('ID'))
>>> x
[<frame ID=200 name=Criminal_process>,
<frame ID=500 name=Criminal_investigation>,
<frame ID=692 name=Crime_scenario>,
<frame ID=700 name=Committing_crime>]
A brief intro to Frames (excerpted from "FrameNet II: Extended
Theory and Practice" by Ruppenhofer et. al., 2010):
A Frame is a script-like conceptual structure that describes a
particular type of situation, object, or event along with the
participants and props that are needed for that Frame. For
example, the "Apply_heat" frame describes a common situation
involving a Cook, some Food, and a Heating_Instrument, and is
evoked by words such as bake, blanch, boil, broil, brown,
simmer, steam, etc.
We call the roles of a Frame "frame elements" (FEs) and the
frame-evoking words are called "lexical units" (LUs).
FrameNet includes relations between Frames. Several types of
relations are defined, of which the most important are:
- Inheritance: An IS-A relation. The child frame is a subtype
of the parent frame, and each FE in the parent is bound to
a corresponding FE in the child. An example is the
"Revenge" frame which inherits from the
"Rewards_and_punishments" frame.
- Using: The child frame presupposes the parent frame as
background, e.g the "Speed" frame "uses" (or presupposes)
the "Motion" frame; however, not all parent FEs need to be
bound to child FEs.
- Subframe: The child frame is a subevent of a complex event
represented by the parent, e.g. the "Criminal_process" frame
has subframes of "Arrest", "Arraignment", "Trial", and
"Sentencing".
- Perspective_on: The child frame provides a particular
perspective on an un-perspectivized parent frame. A pair of
examples consists of the "Hiring" and "Get_a_job" frames,
which perspectivize the "Employment_start" frame from the
Employer's and the Employee's point of view, respectively.
:param name: A regular expression pattern used to match against
Frame names. If 'name' is None, then a list of all
Framenet Frames will be returned.
:type name: str
:return: A list of matching Frames (or all Frames).
:rtype: list(AttrDict)
"""
try:
fIDs = list(self._frame_idx.keys())
except AttributeError:
self._buildframeindex()
fIDs = list(self._frame_idx.keys())
if name is not None:
return PrettyList(
self.frame(fID) for fID, finfo in self.frame_ids_and_names(name).items()
)
else:
return PrettyLazyMap(self.frame, fIDs)
def frame_ids_and_names(self, name=None):
"""
Uses the frame index, which is much faster than looking up each frame definition
if only the names and IDs are needed.
"""
if not self._frame_idx:
self._buildframeindex()
return dict(
(fID, finfo.name)
for fID, finfo in self._frame_idx.items()
if name is None or re.search(name, finfo.name) is not None
)
def fes(self, name=None, frame=None):
"""
Lists frame element objects. If 'name' is provided, this is treated as
a case-insensitive regular expression to filter by frame name.
(Case-insensitivity is because casing of frame element names is not always
consistent across frames.) Specify 'frame' to filter by a frame name pattern,
ID, or object.
>>> from nltk.corpus import framenet as fn
>>> fn.fes('Noise_maker')
[<fe ID=6043 name=Noise_maker>]
>>> sorted([(fe.frame.name,fe.name) for fe in fn.fes('sound')])
[('Cause_to_make_noise', 'Sound_maker'), ('Make_noise', 'Sound'),
('Make_noise', 'Sound_source'), ('Sound_movement', 'Location_of_sound_source'),
('Sound_movement', 'Sound'), ('Sound_movement', 'Sound_source'),
('Sounds', 'Component_sound'), ('Sounds', 'Location_of_sound_source'),
('Sounds', 'Sound_source'), ('Vocalizations', 'Location_of_sound_source'),
('Vocalizations', 'Sound_source')]
>>> sorted([(fe.frame.name,fe.name) for fe in fn.fes('sound',r'(?i)make_noise')])
[('Cause_to_make_noise', 'Sound_maker'),
('Make_noise', 'Sound'),
('Make_noise', 'Sound_source')]
>>> sorted(set(fe.name for fe in fn.fes('^sound')))
['Sound', 'Sound_maker', 'Sound_source']
>>> len(fn.fes('^sound$'))
2
:param name: A regular expression pattern used to match against
frame element names. If 'name' is None, then a list of all
frame elements will be returned.
:type name: str
:return: A list of matching frame elements
:rtype: list(AttrDict)
"""
# what frames are we searching in?
if frame is not None:
if isinstance(frame, int):
frames = [self.frame(frame)]
elif isinstance(frame, str):
frames = self.frames(frame)
else:
frames = [frame]
else:
frames = self.frames()
return PrettyList(
fe
for f in frames
for fename, fe in f.FE.items()
if name is None or re.search(name, fename, re.I)
)
def lus(self, name=None, frame=None):
"""
Obtain details for lexical units.
Optionally restrict by lexical unit name pattern, and/or to a certain frame
or frames whose name matches a pattern.
>>> from nltk.corpus import framenet as fn
>>> len(fn.lus()) in (11829, 13572) # FN 1.5 and 1.7, resp.
True
>>> PrettyList(sorted(fn.lus(r'(?i)a little'), key=itemgetter('ID')), maxReprSize=0, breakLines=True)
[<lu ID=14733 name=a little.n>,
<lu ID=14743 name=a little.adv>,
<lu ID=14744 name=a little bit.adv>]
>>> PrettyList(sorted(fn.lus(r'interest', r'(?i)stimulus'), key=itemgetter('ID')))
[<lu ID=14894 name=interested.a>, <lu ID=14920 name=interesting.a>]
A brief intro to Lexical Units (excerpted from "FrameNet II:
Extended Theory and Practice" by Ruppenhofer et. al., 2010):
A lexical unit (LU) is a pairing of a word with a meaning. For
example, the "Apply_heat" Frame describes a common situation
involving a Cook, some Food, and a Heating Instrument, and is
_evoked_ by words such as bake, blanch, boil, broil, brown,
simmer, steam, etc. These frame-evoking words are the LUs in the
Apply_heat frame. Each sense of a polysemous word is a different
LU.
We have used the word "word" in talking about LUs. The reality
is actually rather complex. When we say that the word "bake" is
polysemous, we mean that the lemma "bake.v" (which has the
word-forms "bake", "bakes", "baked", and "baking") is linked to
three different frames:
- Apply_heat: "Michelle baked the potatoes for 45 minutes."
- Cooking_creation: "Michelle baked her mother a cake for her birthday."
- Absorb_heat: "The potatoes have to bake for more than 30 minutes."
These constitute three different LUs, with different
definitions.
Multiword expressions such as "given name" and hyphenated words
like "shut-eye" can also be LUs. Idiomatic phrases such as
"middle of nowhere" and "give the slip (to)" are also defined as
LUs in the appropriate frames ("Isolated_places" and "Evading",
respectively), and their internal structure is not analyzed.
Framenet provides multiple annotated examples of each sense of a
word (i.e. each LU). Moreover, the set of examples
(approximately 20 per LU) illustrates all of the combinatorial
possibilities of the lexical unit.
Each LU is linked to a Frame, and hence to the other words which
evoke that Frame. This makes the FrameNet database similar to a
thesaurus, grouping together semantically similar words.
In the simplest case, frame-evoking words are verbs such as
"fried" in:
"Matilde fried the catfish in a heavy iron skillet."
Sometimes event nouns may evoke a Frame. For example,
"reduction" evokes "Cause_change_of_scalar_position" in:
"...the reduction of debt levels to $665 million from $2.6 billion."
Adjectives may also evoke a Frame. For example, "asleep" may
evoke the "Sleep" frame as in:
"They were asleep for hours."
Many common nouns, such as artifacts like "hat" or "tower",
typically serve as dependents rather than clearly evoking their
own frames.
:param name: A regular expression pattern used to search the LU
names. Note that LU names take the form of a dotted
string (e.g. "run.v" or "a little.adv") in which a
lemma preceeds the "." and a POS follows the
dot. The lemma may be composed of a single lexeme
(e.g. "run") or of multiple lexemes (e.g. "a
little"). If 'name' is not given, then all LUs will
be returned.
The valid POSes are:
v - verb
n - noun
a - adjective
adv - adverb
prep - preposition
num - numbers
intj - interjection
art - article
c - conjunction
scon - subordinating conjunction
:type name: str
:type frame: str or int or frame
:return: A list of selected (or all) lexical units
:rtype: list of LU objects (dicts). See the lu() function for info
about the specifics of LU objects.
"""
if not self._lu_idx:
self._buildluindex()
if name is not None: # match LUs, then restrict by frame
result = PrettyList(
self.lu(luID) for luID, luName in self.lu_ids_and_names(name).items()
)
if frame is not None:
if isinstance(frame, int):
frameIDs = {frame}
elif isinstance(frame, str):
frameIDs = {f.ID for f in self.frames(frame)}
else:
frameIDs = {frame.ID}
result = PrettyList(lu for lu in result if lu.frame.ID in frameIDs)
elif frame is not None: # all LUs in matching frames
if isinstance(frame, int):
frames = [self.frame(frame)]
elif isinstance(frame, str):
frames = self.frames(frame)
else:
frames = [frame]
result = PrettyLazyIteratorList(
iter(LazyConcatenation(list(f.lexUnit.values()) for f in frames))
)
else: # all LUs
luIDs = [
luID
for luID, lu in self._lu_idx.items()
if lu.status not in self._bad_statuses
]
result = PrettyLazyMap(self.lu, luIDs)
return result
def lu_ids_and_names(self, name=None):
"""
Uses the LU index, which is much faster than looking up each LU definition
if only the names and IDs are needed.
"""
if not self._lu_idx:
self._buildluindex()
return {
luID: luinfo.name
for luID, luinfo in self._lu_idx.items()
if luinfo.status not in self._bad_statuses
and (name is None or re.search(name, luinfo.name) is not None)
}
def docs_metadata(self, name=None):
"""
Return an index of the annotated documents in Framenet.
Details for a specific annotated document can be obtained using this
class's doc() function and pass it the value of the 'ID' field.
>>> from nltk.corpus import framenet as fn
>>> len(fn.docs()) in (78, 107) # FN 1.5 and 1.7, resp.
True
>>> set([x.corpname for x in fn.docs_metadata()])>=set(['ANC', 'KBEval', \
'LUCorpus-v0.3', 'Miscellaneous', 'NTI', 'PropBank'])
True
:param name: A regular expression pattern used to search the
file name of each annotated document. The document's
file name contains the name of the corpus that the
document is from, followed by two underscores "__"
followed by the document name. So, for example, the
file name "LUCorpus-v0.3__20000410_nyt-NEW.xml" is
from the corpus named "LUCorpus-v0.3" and the
document name is "20000410_nyt-NEW.xml".
:type name: str
:return: A list of selected (or all) annotated documents
:rtype: list of dicts, where each dict object contains the following
keys:
- 'name'
- 'ID'
- 'corpid'
- 'corpname'
- 'description'
- 'filename'
"""
try:
ftlist = PrettyList(self._fulltext_idx.values())
except AttributeError:
self._buildcorpusindex()
ftlist = PrettyList(self._fulltext_idx.values())
if name is None:
return ftlist
else:
return PrettyList(
x for x in ftlist if re.search(name, x["filename"]) is not None
)
def docs(self, name=None):
"""
Return a list of the annotated full-text documents in FrameNet,
optionally filtered by a regex to be matched against the document name.
"""
return PrettyLazyMap((lambda x: self.doc(x.ID)), self.docs_metadata(name))
def sents(self, exemplars=True, full_text=True):
"""
Annotated sentences matching the specified criteria.
"""
if exemplars:
if full_text:
return self.exemplars() + self.ft_sents()
else:
return self.exemplars()
elif full_text:
return self.ft_sents()
def annotations(self, luNamePattern=None, exemplars=True, full_text=True):
"""
Frame annotation sets matching the specified criteria.
"""
if exemplars:
epart = PrettyLazyIteratorList(
sent.frameAnnotation for sent in self.exemplars(luNamePattern)
)
else:
epart = []
if full_text:
if luNamePattern is not None:
matchedLUIDs = set(self.lu_ids_and_names(luNamePattern).keys())
ftpart = PrettyLazyIteratorList(
aset
for sent in self.ft_sents()
for aset in sent.annotationSet[1:]
if luNamePattern is None or aset.get("luID", "CXN_ASET") in matchedLUIDs
)
else:
ftpart = []
if exemplars:
if full_text:
return epart + ftpart
else:
return epart
elif full_text:
return ftpart
def exemplars(self, luNamePattern=None, frame=None, fe=None, fe2=None):
"""
Lexicographic exemplar sentences, optionally filtered by LU name and/or 1-2 FEs that
are realized overtly. 'frame' may be a name pattern, frame ID, or frame instance.
'fe' may be a name pattern or FE instance; if specified, 'fe2' may also
be specified to retrieve sentences with both overt FEs (in either order).
"""
if fe is None and fe2 is not None:
raise FramenetError("exemplars(..., fe=None, fe2=<value>) is not allowed")
elif fe is not None and fe2 is not None:
if not isinstance(fe2, str):
if isinstance(fe, str):
# fe2 is specific to a particular frame. swap fe and fe2 so fe is always used to determine the frame.
fe, fe2 = fe2, fe
elif fe.frame is not fe2.frame: # ensure frames match
raise FramenetError(
"exemplars() call with inconsistent `fe` and `fe2` specification (frames must match)"
)
if frame is None and fe is not None and not isinstance(fe, str):
frame = fe.frame
# narrow down to frames matching criteria
lusByFrame = defaultdict(
list
) # frame name -> matching LUs, if luNamePattern is specified
if frame is not None or luNamePattern is not None:
if frame is None or isinstance(frame, str):
if luNamePattern is not None:
frames = set()
for lu in self.lus(luNamePattern, frame=frame):
frames.add(lu.frame.ID)
lusByFrame[lu.frame.name].append(lu)
frames = LazyMap(self.frame, list(frames))
else:
frames = self.frames(frame)
else:
if isinstance(frame, int):
frames = [self.frame(frame)]
else: # frame object
frames = [frame]
if luNamePattern is not None:
lusByFrame = {frame.name: self.lus(luNamePattern, frame=frame)}
if fe is not None: # narrow to frames that define this FE
if isinstance(fe, str):
frames = PrettyLazyIteratorList(
f
for f in frames
if fe in f.FE
or any(re.search(fe, ffe, re.I) for ffe in f.FE.keys())
)
else:
if fe.frame not in frames:
raise FramenetError(
"exemplars() call with inconsistent `frame` and `fe` specification"
)
frames = [fe.frame]
if fe2 is not None: # narrow to frames that ALSO define this FE
if isinstance(fe2, str):
frames = PrettyLazyIteratorList(
f
for f in frames
if fe2 in f.FE
or any(re.search(fe2, ffe, re.I) for ffe in f.FE.keys())
)
# else we already narrowed it to a single frame
else: # frame, luNamePattern are None. fe, fe2 are None or strings
if fe is not None:
frames = {ffe.frame.ID for ffe in self.fes(fe)}
if fe2 is not None:
frames2 = {ffe.frame.ID for ffe in self.fes(fe2)}
frames = frames & frames2
frames = LazyMap(self.frame, list(frames))
else:
frames = self.frames()
# we've narrowed down 'frames'
# now get exemplars for relevant LUs in those frames
def _matching_exs():
for f in frames:
fes = fes2 = None # FEs of interest
if fe is not None:
fes = (
{ffe for ffe in f.FE.keys() if re.search(fe, ffe, re.I)}
if isinstance(fe, str)
else {fe.name}
)
if fe2 is not None:
fes2 = (
{ffe for ffe in f.FE.keys() if re.search(fe2, ffe, re.I)}
if isinstance(fe2, str)
else {fe2.name}
)
for lu in (
lusByFrame[f.name]
if luNamePattern is not None
else f.lexUnit.values()
):
for ex in lu.exemplars:
if (fes is None or self._exemplar_of_fes(ex, fes)) and (
fes2 is None or self._exemplar_of_fes(ex, fes2)
):
yield ex
return PrettyLazyIteratorList(_matching_exs())
def _exemplar_of_fes(self, ex, fes=None):
"""
Given an exemplar sentence and a set of FE names, return the subset of FE names
that are realized overtly in the sentence on the FE, FE2, or FE3 layer.
If 'fes' is None, returns all overt FE names.
"""
overtNames = set(list(zip(*ex.FE[0]))[2]) if ex.FE[0] else set()
if "FE2" in ex:
overtNames |= set(list(zip(*ex.FE2[0]))[2]) if ex.FE2[0] else set()
if "FE3" in ex:
overtNames |= set(list(zip(*ex.FE3[0]))[2]) if ex.FE3[0] else set()
return overtNames & fes if fes is not None else overtNames
def ft_sents(self, docNamePattern=None):
"""
Full-text annotation sentences, optionally filtered by document name.
"""
return PrettyLazyIteratorList(
sent for d in self.docs(docNamePattern) for sent in d.sentence
)
def frame_relation_types(self):
"""
Obtain a list of frame relation types.
>>> from nltk.corpus import framenet as fn
>>> frts = sorted(fn.frame_relation_types(), key=itemgetter('ID'))
>>> isinstance(frts, list)
True
>>> len(frts) in (9, 10) # FN 1.5 and 1.7, resp.
True
>>> PrettyDict(frts[0], breakLines=True)
{'ID': 1,
'_type': 'framerelationtype',
'frameRelations': [<Parent=Event -- Inheritance -> Child=Change_of_consistency>, <Parent=Event -- Inheritance -> Child=Rotting>, ...],
'name': 'Inheritance',
'subFrameName': 'Child',
'superFrameName': 'Parent'}
:return: A list of all of the frame relation types in framenet
:rtype: list(dict)
"""
if not self._freltyp_idx:
self._buildrelationindex()
return self._freltyp_idx.values()
def frame_relations(self, frame=None, frame2=None, type=None):
"""
:param frame: (optional) frame object, name, or ID; only relations involving
this frame will be returned
:param frame2: (optional; 'frame' must be a different frame) only show relations
between the two specified frames, in either direction
:param type: (optional) frame relation type (name or object); show only relations
of this type
:type frame: int or str or AttrDict
:return: A list of all of the frame relations in framenet
:rtype: list(dict)
>>> from nltk.corpus import framenet as fn
>>> frels = fn.frame_relations()
>>> isinstance(frels, list)
True
>>> len(frels) in (1676, 2070) # FN 1.5 and 1.7, resp.
True
>>> PrettyList(fn.frame_relations('Cooking_creation'), maxReprSize=0, breakLines=True)
[<Parent=Intentionally_create -- Inheritance -> Child=Cooking_creation>,
<Parent=Apply_heat -- Using -> Child=Cooking_creation>,
<MainEntry=Apply_heat -- See_also -> ReferringEntry=Cooking_creation>]
>>> PrettyList(fn.frame_relations(274), breakLines=True)
[<Parent=Avoiding -- Inheritance -> Child=Dodging>,
<Parent=Avoiding -- Inheritance -> Child=Evading>, ...]
>>> PrettyList(fn.frame_relations(fn.frame('Cooking_creation')), breakLines=True)
[<Parent=Intentionally_create -- Inheritance -> Child=Cooking_creation>,
<Parent=Apply_heat -- Using -> Child=Cooking_creation>, ...]
>>> PrettyList(fn.frame_relations('Cooking_creation', type='Inheritance'))
[<Parent=Intentionally_create -- Inheritance -> Child=Cooking_creation>]
>>> PrettyList(fn.frame_relations('Cooking_creation', 'Apply_heat'), breakLines=True)
[<Parent=Apply_heat -- Using -> Child=Cooking_creation>,
<MainEntry=Apply_heat -- See_also -> ReferringEntry=Cooking_creation>]
"""
relation_type = type
if not self._frel_idx:
self._buildrelationindex()
rels = None
if relation_type is not None:
if not isinstance(relation_type, dict):
type = [rt for rt in self.frame_relation_types() if rt.name == type][0]
assert isinstance(type, dict)
# lookup by 'frame'
if frame is not None:
if isinstance(frame, dict) and "frameRelations" in frame:
rels = PrettyList(frame.frameRelations)
else:
if not isinstance(frame, int):
if isinstance(frame, dict):
frame = frame.ID
else:
frame = self.frame_by_name(frame).ID
rels = [self._frel_idx[frelID] for frelID in self._frel_f_idx[frame]]
# filter by 'type'
if type is not None:
rels = [rel for rel in rels if rel.type is type]
elif type is not None:
# lookup by 'type'
rels = type.frameRelations
else:
rels = self._frel_idx.values()
# filter by 'frame2'
if frame2 is not None:
if frame is None:
raise FramenetError(
"frame_relations(frame=None, frame2=<value>) is not allowed"
)
if not isinstance(frame2, int):
if isinstance(frame2, dict):
frame2 = frame2.ID
else:
frame2 = self.frame_by_name(frame2).ID
if frame == frame2:
raise FramenetError(
"The two frame arguments to frame_relations() must be different frames"
)
rels = [
rel
for rel in rels
if rel.superFrame.ID == frame2 or rel.subFrame.ID == frame2
]
return PrettyList(
sorted(
rels,
key=lambda frel: (frel.type.ID, frel.superFrameName, frel.subFrameName),
)
)
def fe_relations(self):
"""
Obtain a list of frame element relations.
>>> from nltk.corpus import framenet as fn
>>> ferels = fn.fe_relations()
>>> isinstance(ferels, list)
True
>>> len(ferels) in (10020, 12393) # FN 1.5 and 1.7, resp.
True
>>> PrettyDict(ferels[0], breakLines=True)
{'ID': 14642,
'_type': 'ferelation',
'frameRelation': <Parent=Abounding_with -- Inheritance -> Child=Lively_place>,
'subFE': <fe ID=11370 name=Degree>,
'subFEName': 'Degree',
'subFrame': <frame ID=1904 name=Lively_place>,
'subID': 11370,
'supID': 2271,
'superFE': <fe ID=2271 name=Degree>,
'superFEName': 'Degree',
'superFrame': <frame ID=262 name=Abounding_with>,
'type': <framerelationtype ID=1 name=Inheritance>}
:return: A list of all of the frame element relations in framenet
:rtype: list(dict)
"""
if not self._ferel_idx:
self._buildrelationindex()
return PrettyList(
sorted(
self._ferel_idx.values(),
key=lambda ferel: (
ferel.type.ID,
ferel.frameRelation.superFrameName,
ferel.superFEName,
ferel.frameRelation.subFrameName,
ferel.subFEName,
),
)
)
def semtypes(self):
"""
Obtain a list of semantic types.
>>> from nltk.corpus import framenet as fn
>>> stypes = fn.semtypes()
>>> len(stypes) in (73, 109) # FN 1.5 and 1.7, resp.
True
>>> sorted(stypes[0].keys())
['ID', '_type', 'abbrev', 'definition', 'definitionMarkup', 'name', 'rootType', 'subTypes', 'superType']
:return: A list of all of the semantic types in framenet
:rtype: list(dict)
"""
if not self._semtypes:
self._loadsemtypes()
return PrettyList(
self._semtypes[i] for i in self._semtypes if isinstance(i, int)
)
def _load_xml_attributes(self, d, elt):
"""
Extracts a subset of the attributes from the given element and
returns them in a dictionary.
:param d: A dictionary in which to store the attributes.
:type d: dict
:param elt: An ElementTree Element
:type elt: Element
:return: Returns the input dict ``d`` possibly including attributes from ``elt``
:rtype: dict
"""
d = type(d)(d)
try:
attr_dict = elt.attrib
except AttributeError:
return d
if attr_dict is None:
return d
# Ignore these attributes when loading attributes from an xml node
ignore_attrs = [ #'cBy', 'cDate', 'mDate', # <-- annotation metadata that could be of interest
"xsi",
"schemaLocation",
"xmlns",
"bgColor",
"fgColor",
]
for attr in attr_dict:
if any(attr.endswith(x) for x in ignore_attrs):
continue
val = attr_dict[attr]
if val.isdigit():
d[attr] = int(val)
else:
d[attr] = val
return d
def _strip_tags(self, data):
"""
Gets rid of all tags and newline characters from the given input
:return: A cleaned-up version of the input string
:rtype: str
"""
try:
"""
# Look for boundary issues in markup. (Sometimes FEs are pluralized in definitions.)
m = re.search(r'\w[<][^/]|[<][/][^>]+[>](s\w|[a-rt-z0-9])', data)
if m:
print('Markup boundary:', data[max(0,m.start(0)-10):m.end(0)+10].replace('\n',' '), file=sys.stderr)
"""
data = data.replace("<t>", "")
data = data.replace("</t>", "")
data = re.sub('<fex name="[^"]+">', "", data)
data = data.replace("</fex>", "")
data = data.replace("<fen>", "")
data = data.replace("</fen>", "")
data = data.replace("<m>", "")
data = data.replace("</m>", "")
data = data.replace("<ment>", "")
data = data.replace("</ment>", "")
data = data.replace("<ex>", "'")
data = data.replace("</ex>", "'")
data = data.replace("<gov>", "")
data = data.replace("</gov>", "")
data = data.replace("<x>", "")
data = data.replace("</x>", "")
# Get rid of <def-root> and </def-root> tags
data = data.replace("<def-root>", "")
data = data.replace("</def-root>", "")
data = data.replace("\n", " ")
except AttributeError:
pass
return data
def _handle_elt(self, elt, tagspec=None):
"""Extracts and returns the attributes of the given element"""
return self._load_xml_attributes(AttrDict(), elt)
def _handle_fulltextindex_elt(self, elt, tagspec=None):
"""
Extracts corpus/document info from the fulltextIndex.xml file.
Note that this function "flattens" the information contained
in each of the "corpus" elements, so that each "document"
element will contain attributes for the corpus and
corpusid. Also, each of the "document" items will contain a
new attribute called "filename" that is the base file name of
the xml file for the document in the "fulltext" subdir of the
Framenet corpus.
"""
ftinfo = self._load_xml_attributes(AttrDict(), elt)
corpname = ftinfo.name
corpid = ftinfo.ID
retlist = []
for sub in elt:
if sub.tag.endswith("document"):
doc = self._load_xml_attributes(AttrDict(), sub)
if "name" in doc:
docname = doc.name
else:
docname = doc.description
doc.filename = "{0}__{1}.xml".format(corpname, docname)
doc.URL = (
self._fnweb_url + "/" + self._fulltext_dir + "/" + doc.filename
)
doc.corpname = corpname
doc.corpid = corpid
retlist.append(doc)
return retlist
def _handle_frame_elt(self, elt, ignorekeys=[]):
"""Load the info for a Frame from a frame xml file"""
frinfo = self._load_xml_attributes(AttrDict(), elt)
frinfo["_type"] = "frame"
frinfo["definition"] = ""
frinfo["definitionMarkup"] = ""
frinfo["FE"] = PrettyDict()
frinfo["FEcoreSets"] = []
frinfo["lexUnit"] = PrettyDict()
frinfo["semTypes"] = []
for k in ignorekeys:
if k in frinfo:
del frinfo[k]
for sub in elt:
if sub.tag.endswith("definition") and "definition" not in ignorekeys:
frinfo["definitionMarkup"] = sub.text
frinfo["definition"] = self._strip_tags(sub.text)
elif sub.tag.endswith("FE") and "FE" not in ignorekeys:
feinfo = self._handle_fe_elt(sub)
frinfo["FE"][feinfo.name] = feinfo
feinfo["frame"] = frinfo # backpointer
elif sub.tag.endswith("FEcoreSet") and "FEcoreSet" not in ignorekeys:
coreset = self._handle_fecoreset_elt(sub)
# assumes all FEs have been loaded before coresets
frinfo["FEcoreSets"].append(
PrettyList(frinfo["FE"][fe.name] for fe in coreset)
)
elif sub.tag.endswith("lexUnit") and "lexUnit" not in ignorekeys:
luentry = self._handle_framelexunit_elt(sub)
if luentry["status"] in self._bad_statuses:
# problematic LU entry; ignore it
continue
luentry["frame"] = frinfo
luentry["URL"] = (
self._fnweb_url
+ "/"
+ self._lu_dir
+ "/"
+ "lu{0}.xml".format(luentry["ID"])
)
luentry["subCorpus"] = Future(
(lambda lu: lambda: self._lu_file(lu).subCorpus)(luentry)
)
luentry["exemplars"] = Future(
(lambda lu: lambda: self._lu_file(lu).exemplars)(luentry)
)
frinfo["lexUnit"][luentry.name] = luentry
if not self._lu_idx:
self._buildluindex()
self._lu_idx[luentry.ID] = luentry
elif sub.tag.endswith("semType") and "semTypes" not in ignorekeys:
semtypeinfo = self._load_xml_attributes(AttrDict(), sub)
frinfo["semTypes"].append(self.semtype(semtypeinfo.ID))
frinfo["frameRelations"] = self.frame_relations(frame=frinfo)
# resolve 'requires' and 'excludes' links between FEs of this frame
for fe in frinfo.FE.values():
if fe.requiresFE:
name, ID = fe.requiresFE.name, fe.requiresFE.ID
fe.requiresFE = frinfo.FE[name]
assert fe.requiresFE.ID == ID
if fe.excludesFE:
name, ID = fe.excludesFE.name, fe.excludesFE.ID
fe.excludesFE = frinfo.FE[name]
assert fe.excludesFE.ID == ID
return frinfo
def _handle_fecoreset_elt(self, elt):
"""Load fe coreset info from xml."""
info = self._load_xml_attributes(AttrDict(), elt)
tmp = []
for sub in elt:
tmp.append(self._load_xml_attributes(AttrDict(), sub))
return tmp
def _handle_framerelationtype_elt(self, elt, *args):
"""Load frame-relation element and its child fe-relation elements from frRelation.xml."""
info = self._load_xml_attributes(AttrDict(), elt)
info["_type"] = "framerelationtype"
info["frameRelations"] = PrettyList()
for sub in elt:
if sub.tag.endswith("frameRelation"):
frel = self._handle_framerelation_elt(sub)
frel["type"] = info # backpointer
for ferel in frel.feRelations:
ferel["type"] = info
info["frameRelations"].append(frel)
return info
def _handle_framerelation_elt(self, elt):
"""Load frame-relation element and its child fe-relation elements from frRelation.xml."""
info = self._load_xml_attributes(AttrDict(), elt)
assert info["superFrameName"] != info["subFrameName"], (elt, info)
info["_type"] = "framerelation"
info["feRelations"] = PrettyList()
for sub in elt:
if sub.tag.endswith("FERelation"):
ferel = self._handle_elt(sub)
ferel["_type"] = "ferelation"
ferel["frameRelation"] = info # backpointer
info["feRelations"].append(ferel)
return info
def _handle_fulltextannotation_elt(self, elt):
"""Load full annotation info for a document from its xml
file. The main element (fullTextAnnotation) contains a 'header'
element (which we ignore here) and a bunch of 'sentence'
elements."""
info = AttrDict()
info["_type"] = "fulltext_annotation"
info["sentence"] = []
for sub in elt:
if sub.tag.endswith("header"):
continue # not used
elif sub.tag.endswith("sentence"):
try:
s = self._handle_fulltext_sentence_elt(sub)
s.doc = info
info["sentence"].append(s)
except FramenetError as e:
print(e)
return info
def _handle_fulltext_sentence_elt(self, elt):
"""Load information from the given 'sentence' element. Each
'sentence' element contains a "text" and "annotationSet" sub
elements."""
info = self._load_xml_attributes(AttrDict(), elt)
info["_type"] = "fulltext_sentence"
info["annotationSet"] = []
info["targets"] = []
target_spans = set()
info["_ascii"] = types.MethodType(
_annotation_ascii, info
) # attach a method for this instance
info["text"] = ""
for sub in elt:
if sub.tag.endswith("text"):
info["text"] = self._strip_tags(sub.text)
elif sub.tag.endswith("annotationSet"):
a = self._handle_fulltextannotationset_elt(
sub, is_pos=(len(info["annotationSet"]) == 0)
)
if "cxnID" in a: # ignoring construction annotations for now
continue
a.sent = info
a.text = info.text
info["annotationSet"].append(a)
if "Target" in a:
for tspan in a.Target:
if tspan in target_spans:
self._warn(
'Duplicate target span "{0}"'.format(
info.text[slice(*tspan)]
),
tspan,
"in sentence",
info["ID"],
info.text,
)
# this can happen in cases like "chemical and biological weapons"
# being annotated as "chemical weapons" and "biological weapons"
else:
target_spans.add(tspan)
info["targets"].append((a.Target, a.luName, a.frameName))
assert info["annotationSet"][0].status == "UNANN"
info["POS"] = info["annotationSet"][0].POS
info["POS_tagset"] = info["annotationSet"][0].POS_tagset
return info
def _handle_fulltextannotationset_elt(self, elt, is_pos=False):
"""Load information from the given 'annotationSet' element. Each
'annotationSet' contains several "layer" elements."""
info = self._handle_luannotationset_elt(elt, is_pos=is_pos)
if not is_pos:
info["_type"] = "fulltext_annotationset"
if "cxnID" not in info: # ignoring construction annotations for now
info["LU"] = self.lu(
info.luID,
luName=info.luName,
frameID=info.frameID,
frameName=info.frameName,
)
info["frame"] = info.LU.frame
return info
def _handle_fulltextlayer_elt(self, elt):
"""Load information from the given 'layer' element. Each
'layer' contains several "label" elements."""
info = self._load_xml_attributes(AttrDict(), elt)
info["_type"] = "layer"
info["label"] = []
for sub in elt:
if sub.tag.endswith("label"):
l = self._load_xml_attributes(AttrDict(), sub)
info["label"].append(l)
return info
def _handle_framelexunit_elt(self, elt):
"""Load the lexical unit info from an xml element in a frame's xml file."""
luinfo = AttrDict()
luinfo["_type"] = "lu"
luinfo = self._load_xml_attributes(luinfo, elt)
luinfo["definition"] = ""
luinfo["definitionMarkup"] = ""
luinfo["sentenceCount"] = PrettyDict()
luinfo["lexemes"] = PrettyList() # multiword LUs have multiple lexemes
luinfo["semTypes"] = PrettyList() # an LU can have multiple semtypes
for sub in elt:
if sub.tag.endswith("definition"):
luinfo["definitionMarkup"] = sub.text
luinfo["definition"] = self._strip_tags(sub.text)
elif sub.tag.endswith("sentenceCount"):
luinfo["sentenceCount"] = self._load_xml_attributes(PrettyDict(), sub)
elif sub.tag.endswith("lexeme"):
lexemeinfo = self._load_xml_attributes(PrettyDict(), sub)
if not isinstance(lexemeinfo.name, str):
# some lexeme names are ints by default: e.g.,
# thousand.num has lexeme with name="1000"
lexemeinfo.name = str(lexemeinfo.name)
luinfo["lexemes"].append(lexemeinfo)
elif sub.tag.endswith("semType"):
semtypeinfo = self._load_xml_attributes(PrettyDict(), sub)
luinfo["semTypes"].append(self.semtype(semtypeinfo.ID))
# sort lexemes by 'order' attribute
# otherwise, e.g., 'write down.v' may have lexemes in wrong order
luinfo["lexemes"].sort(key=lambda x: x.order)
return luinfo
def _handle_lexunit_elt(self, elt, ignorekeys):
"""
Load full info for a lexical unit from its xml file.
This should only be called when accessing corpus annotations
(which are not included in frame files).
"""
luinfo = self._load_xml_attributes(AttrDict(), elt)
luinfo["_type"] = "lu"
luinfo["definition"] = ""
luinfo["definitionMarkup"] = ""
luinfo["subCorpus"] = PrettyList()
luinfo["lexemes"] = PrettyList() # multiword LUs have multiple lexemes
luinfo["semTypes"] = PrettyList() # an LU can have multiple semtypes
for k in ignorekeys:
if k in luinfo:
del luinfo[k]
for sub in elt:
if sub.tag.endswith("header"):
continue # not used
elif sub.tag.endswith("valences"):
continue # not used
elif sub.tag.endswith("definition") and "definition" not in ignorekeys:
luinfo["definitionMarkup"] = sub.text
luinfo["definition"] = self._strip_tags(sub.text)
elif sub.tag.endswith("subCorpus") and "subCorpus" not in ignorekeys:
sc = self._handle_lusubcorpus_elt(sub)
if sc is not None:
luinfo["subCorpus"].append(sc)
elif sub.tag.endswith("lexeme") and "lexeme" not in ignorekeys:
luinfo["lexemes"].append(self._load_xml_attributes(PrettyDict(), sub))
elif sub.tag.endswith("semType") and "semType" not in ignorekeys:
semtypeinfo = self._load_xml_attributes(AttrDict(), sub)
luinfo["semTypes"].append(self.semtype(semtypeinfo.ID))
return luinfo
def _handle_lusubcorpus_elt(self, elt):
"""Load a subcorpus of a lexical unit from the given xml."""
sc = AttrDict()
try:
sc["name"] = elt.get("name")
except AttributeError:
return None
sc["_type"] = "lusubcorpus"
sc["sentence"] = []
for sub in elt:
if sub.tag.endswith("sentence"):
s = self._handle_lusentence_elt(sub)
if s is not None:
sc["sentence"].append(s)
return sc
def _handle_lusentence_elt(self, elt):
"""Load a sentence from a subcorpus of an LU from xml."""
info = self._load_xml_attributes(AttrDict(), elt)
info["_type"] = "lusentence"
info["annotationSet"] = []
info["_ascii"] = types.MethodType(
_annotation_ascii, info
) # attach a method for this instance
for sub in elt:
if sub.tag.endswith("text"):
info["text"] = self._strip_tags(sub.text)
elif sub.tag.endswith("annotationSet"):
annset = self._handle_luannotationset_elt(
sub, is_pos=(len(info["annotationSet"]) == 0)
)
if annset is not None:
assert annset.status == "UNANN" or "FE" in annset, annset
if annset.status != "UNANN":
info["frameAnnotation"] = annset
# copy layer info up to current level
for k in (
"Target",
"FE",
"FE2",
"FE3",
"GF",
"PT",
"POS",
"POS_tagset",
"Other",
"Sent",
"Verb",
"Noun",
"Adj",
"Adv",
"Prep",
"Scon",
"Art",
):
if k in annset:
info[k] = annset[k]
info["annotationSet"].append(annset)
annset["sent"] = info
annset["text"] = info.text
return info
def _handle_luannotationset_elt(self, elt, is_pos=False):
"""Load an annotation set from a sentence in an subcorpus of an LU"""
info = self._load_xml_attributes(AttrDict(), elt)
info["_type"] = "posannotationset" if is_pos else "luannotationset"
info["layer"] = []
info["_ascii"] = types.MethodType(
_annotation_ascii, info
) # attach a method for this instance
if "cxnID" in info: # ignoring construction annotations for now.
return info
for sub in elt:
if sub.tag.endswith("layer"):
l = self._handle_lulayer_elt(sub)
if l is not None:
overt = []
ni = {} # null instantiations
info["layer"].append(l)
for lbl in l.label:
if "start" in lbl:
thespan = (lbl.start, lbl.end + 1, lbl.name)
if l.name not in (
"Sent",
"Other",
): # 'Sent' and 'Other' layers sometimes contain accidental duplicate spans
assert thespan not in overt, (info.ID, l.name, thespan)
overt.append(thespan)
else: # null instantiation
if lbl.name in ni:
self._warn(
"FE with multiple NI entries:",
lbl.name,
ni[lbl.name],
lbl.itype,
)
else:
ni[lbl.name] = lbl.itype
overt = sorted(overt)
if l.name == "Target":
if not overt:
self._warn(
"Skipping empty Target layer in annotation set ID={0}".format(
info.ID
)
)
continue
assert all(lblname == "Target" for i, j, lblname in overt)
if "Target" in info:
self._warn(
"Annotation set {0} has multiple Target layers".format(
info.ID
)
)
else:
info["Target"] = [(i, j) for (i, j, _) in overt]
elif l.name == "FE":
if l.rank == 1:
assert "FE" not in info
info["FE"] = (overt, ni)
# assert False,info
else:
# sometimes there are 3 FE layers! e.g. Change_position_on_a_scale.fall.v
assert 2 <= l.rank <= 3, l.rank
k = "FE" + str(l.rank)
assert k not in info
info[k] = (overt, ni)
elif l.name in ("GF", "PT"):
assert l.rank == 1
info[l.name] = overt
elif l.name in ("BNC", "PENN"):
assert l.rank == 1
info["POS"] = overt
info["POS_tagset"] = l.name
else:
if is_pos:
if l.name not in ("NER", "WSL"):
self._warn(
"Unexpected layer in sentence annotationset:",
l.name,
)
else:
if l.name not in (
"Sent",
"Verb",
"Noun",
"Adj",
"Adv",
"Prep",
"Scon",
"Art",
"Other",
):
self._warn(
"Unexpected layer in frame annotationset:", l.name
)
info[l.name] = overt
if not is_pos and "cxnID" not in info:
if "Target" not in info:
self._warn("Missing target in annotation set ID={0}".format(info.ID))
assert "FE" in info
if "FE3" in info:
assert "FE2" in info
return info
def _handle_lulayer_elt(self, elt):
"""Load a layer from an annotation set"""
layer = self._load_xml_attributes(AttrDict(), elt)
layer["_type"] = "lulayer"
layer["label"] = []
for sub in elt:
if sub.tag.endswith("label"):
l = self._load_xml_attributes(AttrDict(), sub)
if l is not None:
layer["label"].append(l)
return layer
def _handle_fe_elt(self, elt):
feinfo = self._load_xml_attributes(AttrDict(), elt)
feinfo["_type"] = "fe"
feinfo["definition"] = ""
feinfo["definitionMarkup"] = ""
feinfo["semType"] = None
feinfo["requiresFE"] = None
feinfo["excludesFE"] = None
for sub in elt:
if sub.tag.endswith("definition"):
feinfo["definitionMarkup"] = sub.text
feinfo["definition"] = self._strip_tags(sub.text)
elif sub.tag.endswith("semType"):
stinfo = self._load_xml_attributes(AttrDict(), sub)
feinfo["semType"] = self.semtype(stinfo.ID)
elif sub.tag.endswith("requiresFE"):
feinfo["requiresFE"] = self._load_xml_attributes(AttrDict(), sub)
elif sub.tag.endswith("excludesFE"):
feinfo["excludesFE"] = self._load_xml_attributes(AttrDict(), sub)
return feinfo
def _handle_semtype_elt(self, elt, tagspec=None):
semt = self._load_xml_attributes(AttrDict(), elt)
semt["_type"] = "semtype"
semt["superType"] = None
semt["subTypes"] = PrettyList()
for sub in elt:
if sub.text is not None:
semt["definitionMarkup"] = sub.text
semt["definition"] = self._strip_tags(sub.text)
else:
supertypeinfo = self._load_xml_attributes(AttrDict(), sub)
semt["superType"] = supertypeinfo
# the supertype may not have been loaded yet
return semt
#
# Demo
#
def demo():
from nltk.corpus import framenet as fn
#
# It is not necessary to explicitly build the indexes by calling
# buildindexes(). We do this here just for demo purposes. If the
# indexes are not built explicitely, they will be built as needed.
#
print("Building the indexes...")
fn.buildindexes()
#
# Get some statistics about the corpus
#
print("Number of Frames:", len(fn.frames()))
print("Number of Lexical Units:", len(fn.lus()))
print("Number of annotated documents:", len(fn.docs()))
print()
#
# Frames
#
print(
'getting frames whose name matches the (case insensitive) regex: "(?i)medical"'
)
medframes = fn.frames(r"(?i)medical")
print('Found {0} Frames whose name matches "(?i)medical":'.format(len(medframes)))
print([(f.name, f.ID) for f in medframes])
#
# store the first frame in the list of frames
#
tmp_id = medframes[0].ID
m_frame = fn.frame(tmp_id) # reads all info for the frame
#
# get the frame relations
#
print(
'\nNumber of frame relations for the "{0}" ({1}) frame:'.format(
m_frame.name, m_frame.ID
),
len(m_frame.frameRelations),
)
for fr in m_frame.frameRelations:
print(" ", fr)
#
# get the names of the Frame Elements
#
print(
'\nNumber of Frame Elements in the "{0}" frame:'.format(m_frame.name),
len(m_frame.FE),
)
print(" ", [x for x in m_frame.FE])
#
# get the names of the "Core" Frame Elements
#
print('\nThe "core" Frame Elements in the "{0}" frame:'.format(m_frame.name))
print(" ", [x.name for x in m_frame.FE.values() if x.coreType == "Core"])
#
# get all of the Lexical Units that are incorporated in the
# 'Ailment' FE of the 'Medical_conditions' frame (id=239)
#
print('\nAll Lexical Units that are incorporated in the "Ailment" FE:')
m_frame = fn.frame(239)
ailment_lus = [
x
for x in m_frame.lexUnit.values()
if "incorporatedFE" in x and x.incorporatedFE == "Ailment"
]
print(" ", [x.name for x in ailment_lus])
#
# get all of the Lexical Units for the frame
#
print(
'\nNumber of Lexical Units in the "{0}" frame:'.format(m_frame.name),
len(m_frame.lexUnit),
)
print(" ", [x.name for x in m_frame.lexUnit.values()][:5], "...")
#
# get basic info on the second LU in the frame
#
tmp_id = m_frame.lexUnit["ailment.n"].ID # grab the id of the specified LU
luinfo = fn.lu_basic(tmp_id) # get basic info on the LU
print("\nInformation on the LU: {0}".format(luinfo.name))
pprint(luinfo)
#
# Get a list of all of the corpora used for fulltext annotation
#
print("\nNames of all of the corpora used for fulltext annotation:")
allcorpora = set(x.corpname for x in fn.docs_metadata())
pprint(list(allcorpora))
#
# Get the names of the annotated documents in the first corpus
#
firstcorp = list(allcorpora)[0]
firstcorp_docs = fn.docs(firstcorp)
print('\nNames of the annotated documents in the "{0}" corpus:'.format(firstcorp))
pprint([x.filename for x in firstcorp_docs])
#
# Search for frames containing LUs whose name attribute matches a
# regexp pattern.
#
# Note: if you were going to be doing a lot of this type of
# searching, you'd want to build an index that maps from
# lemmas to frames because each time frames_by_lemma() is
# called, it has to search through ALL of the frame XML files
# in the db.
print(
'\nSearching for all Frames that have a lemma that matches the regexp: "^run.v$":'
)
pprint(fn.frames_by_lemma(r"^run.v$"))
if __name__ == "__main__":
demo()
framenet15 = LazyCorpusLoader(
"framenet_v15",
FramenetCorpusReader,
[
"frRelation.xml",
"frameIndex.xml",
"fulltextIndex.xml",
"luIndex.xml",
"semTypes.xml",
],
)
framenet = LazyCorpusLoader(
"framenet_v17",
FramenetCorpusReader,
[
"frRelation.xml",
"frameIndex.xml",
"fulltextIndex.xml",
"luIndex.xml",
"semTypes.xml",
],
)
|