Commit Graph

  • 88095666dc * Remove Lexeme struct, preparing to rename Word to Lexeme. Matthew Honnibal 2014-08-24 19:24:42 +0200
  • ce59526011 * Add Word classes Matthew Honnibal 2014-08-24 18:14:08 +0200
  • 3b793cf4f7 * Tests passing for new Word object version Matthew Honnibal 2014-08-24 18:13:53 +0200
  • 9815c7649e * Refactor around Word objects, adapting tests. Tests passing, except for string views. Matthew Honnibal 2014-08-23 19:55:06 +0200
  • 4f01df9152 * Moving to Word objects in place of the Lexeme struct. Matthew Honnibal 2014-08-22 17:32:16 +0200
  • 782806df08 * Moving to Word objects in place of the Lexeme struct. Matthew Honnibal 2014-08-22 17:28:23 +0200
  • 47fbd0475a * Replace the use of dense_hash_map with Python dict Matthew Honnibal 2014-08-22 17:13:09 +0200
  • 6f83dca218 * Fix import for ptb tokenization test Matthew Honnibal 2014-08-22 17:05:44 +0200
  • e289896603 * Fix ptb3 module Matthew Honnibal 2014-08-22 16:35:48 +0200
  • a22101404a * Move en_ptb data Matthew Honnibal 2014-08-22 04:28:51 +0200
  • 89d6faa9c9 * Move en_ptb to ptb3 Matthew Honnibal 2014-08-22 04:23:24 +0200
  • 4bcdd6d31c * Further improvements to spacy docs, tweaks to code. Matthew Honnibal 2014-08-22 04:20:24 +0200
  • 4eb9c2b30f * Add overview doc Matthew Honnibal 2014-08-22 03:38:05 +0200
  • 07ecf5d2f4 * Fixed group_by, removed idea of general attr_of function. Matthew Honnibal 2014-08-22 00:02:37 +0200
  • 811b7a6b91 * Struggling with arbitrary attr access... Matthew Honnibal 2014-08-21 23:49:14 +0200
  • 314658b31c * Improve module docstring Matthew Honnibal 2014-08-21 18:42:47 +0200
  • 8bcd07dbae * More docs work Matthew Honnibal 2014-08-21 17:05:28 +0200
  • d10993f41a * More docs work Matthew Honnibal 2014-08-21 16:37:13 +0200
  • d5403a6fe3 * More docs work Matthew Honnibal 2014-08-21 16:37:06 +0200
  • 248cbb6d07 * Update doc strings Matthew Honnibal 2014-08-21 03:29:15 +0200
  • cbda38e2d9 * Improving docs Matthew Honnibal 2014-08-20 21:09:39 +0200
  • cab7f63fc2 * Temporarily remove sparsehash requirement Matthew Honnibal 2014-08-20 17:12:19 +0200
  • c289867d1f * Add draft sphinx docs files Matthew Honnibal 2014-08-20 17:05:18 +0200
  • bebfd7940d * Upd gitignore Matthew Honnibal 2014-08-20 17:04:33 +0200
  • 76afbd7d69 * Remove compiled orthography file Matthew Honnibal 2014-08-20 17:04:07 +0200
  • f39dcb1d89 * Add orthography Matthew Honnibal 2014-08-20 17:03:44 +0200
  • d42cdbb446 * Compile orthography.latin.pyx Matthew Honnibal 2014-08-20 17:03:19 +0200
  • 0c4c47b074 * Add docs requirements Matthew Honnibal 2014-08-20 17:02:54 +0200
  • bd51742fbd * Remove MurmurHash headers Matthew Honnibal 2014-08-20 17:02:47 +0200
  • 416a324bcf * Add documentation building to fabfile Matthew Honnibal 2014-08-20 17:02:32 +0200
  • a78ad4152d * Broken version being refactored for docs Matthew Honnibal 2014-08-20 13:39:39 +0200
  • 5fddb8d165 * Working refactor, with updated data model for Lexemes Matthew Honnibal 2014-08-19 04:21:20 +0200
  • 3379d7a571 * Reforming data model for lexemes Matthew Honnibal 2014-08-19 02:40:37 +0200
  • e091b6a241 * Base master on temp branch Matthew Honnibal 2014-08-18 23:29:21 +0200
  • 85df22c379 * Remove murmurhash from requirements Matthew Honnibal 2014-08-18 23:26:20 +0200
  • ab9b0daabf * Whitespace Matthew Honnibal 2014-08-18 23:21:49 +0200
  • a2047fa5aa * Add 's suffix to tokenization table Matthew Honnibal 2014-08-18 23:21:37 +0200
  • 1b71cbfe28 * Roll back to using unicode, and never Py_UNICODE. No dependence on murmurhash either. Matthew Honnibal 2014-08-18 20:48:48 +0200
  • bbf9a2c944 * Working version that uses arrays for chunks, which should be more memory efficient Matthew Honnibal 2014-08-18 20:23:54 +0200
  • 8d3f6082be * Working version, adding improvements Matthew Honnibal 2014-08-18 19:59:59 +0200
  • 01469b0888 * Refactor spacy so that chunks return arrays of lexemes, so that there is properly one lexeme per word. Matthew Honnibal 2014-08-18 19:14:00 +0200
  • b94c9b72c9 * WordTree in use. Need to reform the way chunks are handled. Should be properly one Lexeme per word, with split points being the things that are cached. Matthew Honnibal 2014-08-16 20:10:22 +0200
  • 34b68a18ab * Progress to getting WordTree working. Tests pass, but so far it's slower. Matthew Honnibal 2014-08-16 19:59:38 +0200
  • 865cacfaf7 * Remove dependence on murmurhash Matthew Honnibal 2014-08-16 17:37:09 +0200
  • 515d41d325 * Restore string saving to spacy Matthew Honnibal 2014-08-16 16:09:24 +0200
  • 36073b89fe * Restore unicode, work on improving string storage. Matthew Honnibal 2014-08-16 14:35:34 +0200
  • a225ca5b0d * Refactoring tokenizer Matthew Honnibal 2014-08-16 03:22:03 +0200
  • 213a440ffc * Add string decode and encode helpers to string_tools Matthew Honnibal 2014-08-15 23:57:27 +0200
  • 7fd9b2f1f8 * Add murmurhash to setup while we figure out cython includes Matthew Honnibal 2014-08-15 23:56:57 +0200
  • 3ada25b92d * Shifting to WordTree instead of dense_hash_map for storage. Matthew Honnibal 2014-08-15 23:06:46 +0200
  • f11c8e22eb * Remove happax stuff Matthew Honnibal 2014-08-02 22:11:28 +0100
  • 6050209f65 * Update gitignore Matthew Honnibal 2014-08-02 22:06:30 +0100
  • b555e2dc5d * Add hash tests Matthew Honnibal 2014-08-02 21:58:31 +0100
  • d6e07aa922 * Switch to 32bit hash for strings Matthew Honnibal 2014-08-02 21:51:52 +0100
  • 365a2af756 * Restore happax. commit uncommited work Matthew Honnibal 2014-08-02 21:27:03 +0100
  • 6319ff0f22 * Add length property Matthew Honnibal 2014-08-02 21:26:44 +0100
  • 18fb76b2c4 * Removed happax. Not sure if good idea. Matthew Honnibal 2014-08-02 20:53:35 +0100
  • edd38a84b1 * Removing happax stuff. Added length Matthew Honnibal 2014-08-02 20:45:12 +0100
  • fc7c10d7f8 * Ugly but seemingly working fix to the token memory leak Matthew Honnibal 2014-08-01 09:43:19 +0100
  • c7bb6b329c * Don't free clobbered lexemes, as they might be part of a tail Matthew Honnibal 2014-08-01 08:22:38 +0100
  • c48214460e * Free lexemes clobbered as happaxes Matthew Honnibal 2014-08-01 07:40:20 +0100
  • 5b6457e80e * Free lexemes clobbered as happaxes Matthew Honnibal 2014-08-01 07:37:50 +0100
  • d8cb2288ce * Roll back to using murmurhash2 for now Matthew Honnibal 2014-08-01 07:28:47 +0100
  • d4b8bc07ce * Use FixedTable to control index size Matthew Honnibal 2014-08-01 07:27:48 +0100
  • f39211b2b1 * Add FixedTable for hashing Matthew Honnibal 2014-08-01 07:27:21 +0100
  • a44e15f623 * Hack around lack of distribution features for now. Matthew Honnibal 2014-07-31 18:24:51 +0100
  • 4cb88c940b * Fix memory leak in tokenizer, caused by having a fixed vocab. Matthew Honnibal 2014-07-31 18:19:38 +0100
  • 5b81ee716f * Use a sparse_hash_map to store happax vocab items, with a max size. Matthew Honnibal 2014-07-31 17:40:43 +0100
  • a235804730 * Fix setup.py Matthew Honnibal 2014-07-31 02:03:53 +0100
  • 5461399924 * Fix setup.py Matthew Honnibal 2014-07-31 02:03:10 +0100
  • e5da104bc7 * Remove ext files Matthew Honnibal 2014-07-25 16:48:26 +0100
  • b9016c4633 * Switch to using sparsehash and murmurhash libraries out of pip Matthew Honnibal 2014-07-25 15:47:27 +0100
  • 3f7cbb93e0 * Upd requirements Matthew Honnibal 2014-07-23 17:50:34 +0100
  • d2a151ec75 * Push to master Matthew Honnibal 2014-07-23 17:39:02 +0100
  • 150cf6dd3b * Merge string views feature branch Matthew Honnibal 2014-07-23 17:37:48 +0100
  • e80ec5525f * Delete docs attempt Matthew Honnibal 2014-07-23 17:36:37 +0100
  • e494494d80 * Add tests for group_by Matthew Honnibal 2014-07-23 17:36:12 +0100
  • a895fe5ddb * Upd from spacy Matthew Honnibal 2014-07-23 17:35:18 +0100
  • 87bf205b82 * Fix open apostrophe bug Matthew Honnibal 2014-07-07 23:26:01 +0200
  • bc6c1f6156 * Add test for open apostrophe bug Matthew Honnibal 2014-07-07 23:24:20 +0200
  • 571808a274 Group-by seems to be working Matthew Honnibal 2014-07-07 20:27:02 +0200
  • 80b36f9f27 * 710k words per second for counts Matthew Honnibal 2014-07-07 19:12:19 +0200
  • 057c21969b * Refactor for string view features. Working on setting up flags and enums. Matthew Honnibal 2014-07-07 16:58:48 +0200
  • 9fd085bf90 Merge branch 'feature/group_by' into develop Matthew Honnibal 2014-07-07 12:51:40 +0200
  • 1c5ab3b49a * Add tokens module to setup Matthew Honnibal 2014-07-07 12:51:23 +0200
  • f1bcbd4c4e * Reorganized code to accomodate Tokens class. Need string views before group_by and count_by can be done well. Matthew Honnibal 2014-07-07 12:47:21 +0200
  • 6da8010577 Merge branch 'feature/scale' into develop Matthew Honnibal 2014-07-07 08:15:54 +0200
  • 6668e44961 * Whitespace Matthew Honnibal 2014-07-07 08:15:44 +0200
  • 0074ae2fc0 * Switch to dynamically allocating array, based on the document length Matthew Honnibal 2014-07-07 08:05:29 +0200
  • 0575f16ade * Upd requirements Matthew Honnibal 2014-07-07 07:37:29 +0200
  • ff1869ff07 * Fixed major efficiency problem, from not quite grokking pass by reference in cython c++ Matthew Honnibal 2014-07-07 07:36:43 +0200
  • aaae66114c Merge branch 'feature/contractions' into develop Matthew Honnibal 2014-07-07 05:11:43 +0200
  • e60b958b7d * Add test to check how well we match ptb tokenizer. Needs more text. Matthew Honnibal 2014-07-07 05:11:31 +0200
  • 2c431f9fdc * Upd tokenization test Matthew Honnibal 2014-07-07 05:11:04 +0200
  • 0c76143b72 * Give value for assert Matthew Honnibal 2014-07-07 05:10:46 +0200
  • 648d1fe3ed * Compile en_ptb Matthew Honnibal 2014-07-07 05:10:28 +0200
  • 60a2dac13c * Upd gitignore Matthew Honnibal 2014-07-07 05:10:19 +0200
  • e244739dfe * Fix ptb tokenization Matthew Honnibal 2014-07-07 05:10:09 +0200
  • cc3971ce5c * Fix error in tokenization rules Matthew Honnibal 2014-07-07 05:09:34 +0200
  • 997551241f * Upd ptb tokenization rules Matthew Honnibal 2014-07-07 05:09:22 +0200