@@ -36,7 +36,8 @@ static void free_registry_entry(void *key, void *value, void *other) } /* Destroy a Registry whose keys & values have been allocated */ -static void free_registry(Registry r) { +static void free_registry(Registry r) +{ Registry_traverse(r, free_registry_entry, NULL); Registry_destroy(r); } @@ -92,15 +93,18 @@ void tagger_context_add_contextual_rule( TaggerContext *tc, const char *rule ) int tagger_context_add_word_to_wordlist( TaggerContext *tc, const char *word ) { - return Registry_add(tc->wordlist_hash, (VOIDP)strdup(word), (VOIDP)1); + VOIDP pw = (VOIDP)strdup(word); + return Registry_add(tc->wordlist_hash, pw, (VOIDP)1); } int tagger_context_add_goodleft( TaggerContext *tc, const char *word ) { - return Registry_add(tc->good_left_hash, (VOIDP)strdup(word), (VOIDP)1); + VOIDP pw = (VOIDP)strdup(word); + return Registry_add(tc->good_left_hash, pw, (VOIDP)1); } int tagger_context_add_goodright( TaggerContext *tc, const char *word ) { - return Registry_add(tc->good_right_hash, (VOIDP)strdup(word), (VOIDP)1); + VOIDP pw = (VOIDP)strdup(word); + return Registry_add(tc->good_right_hash, pw, (VOIDP)1); } ext/rule_tagger/tagger.c @@ -37,8 +37,9 @@ Although many newly diagnosed patients fear they will not be able to keep workin pairs = tagger.tag( SAMPLE_DOC ) assert_equal [["", ")"], ["", ")"], ["Take", "VB"], ["an", "DT"], ["active", "JJ"], ["role", "NN"], ["in", "IN"], ["your", "PRP$"], ["care", "NN"], ["When", "WRB"], ["it", "PRP"], ["comes", "VBZ"], ["to", "TO"], ["making", "VBG"], ["decisions", "NNS"], ["about", "IN"], ["the", "DT"], ["goals", "NNS"], ["and", "CC"], ["direction", "NN"], ["of", "IN"], ["treatment", "NN"], [",", ","], ["", ")"], ["do", "VBP"], ["", ")"], ["n't", "RB"], ["sit", "VB"], ["back.", "CD"], ["Work", "NN"], ["closely", "RB"], ["and", "CC"], ["actively", "RB"], ["with", "IN"], ["your", "PRP$"], ["oncologist", "NN"], ["and", "CC"], ["the", "DT"], ["rest", "NN"], ["of", "IN"], ["your", "PRP$"], ["medical", "JJ"], ["team.", "NNP"], ["Dont", "NNP"], ["overlook", "VB"], ["clinical", "JJ"], ["trials", "NNS"], ["If", "IN"], ["youre", "NN"], ["eligible", "JJ"], ["to", "TO"], ["enroll", "VB"], ["in", "IN"], ["clinical", "JJ"], ["trials", "NNS"], [",", ","], ["", ")"], ["select", "VB"], ["an", "DT"], ["oncologist", "NN"], ["who", "WP"], ["participates", "VBZ"], ["in", "IN"], ["them.", "JJ"], ["Patients", "NNS"], ["who", "WP"], ["enroll", "VBP"], ["in", "IN"], ["clinical", "JJ"], ["studies", "NNS"], ["receive", "VBP"], ["closer", "JJR"], ["follow-up", "NN"], [",", ","], ["", ")"], ["the", "DT"], ["highest", "JJS"], ["standard-of-care", "JJ"], ["treatment", "NN"], ["and", "CC"], ["access", "NN"], ["to", "TO"], ["experimental", "JJ"], ["therapies", "NNS"], ["at", "IN"], ["no", "DT"], ["extra", "JJ"], ["cost.", "NNP"], ["Maximize", "NNP"], ["your", "PRP$"], ["nutrition", "NN"], ["strategy", "NN"], ["Doing", "NNP"], ["your", "PRP$"], ["best", "JJS"], ["to", "TO"], ["eat", "VB"], ["a", "DT"], ["healthy", "JJ"], [",", ","], ["", ")"], ["well-balanced", "JJ"], ["diet", "NN"], ["is", "VBZ"], ["vital", "JJ"], ["to", "TO"], ["prompt", "VB"], ["healing", "NN"], ["after", "IN"], ["surgery", "NN"], ["and", "CC"], ["for", "IN"], ["recovery", "NN"], ["from", "IN"], ["radiation", "NN"], ["or", "CC"], ["chemotherapy.", "JJ"], ["Many", "JJ"], ["oncology", "NN"], ["practices", "NNS"], ["employ", "VBP"], ["registered", "VBN"], ["dieticians", "NNS"], ["who", "WP"], ["can", "MD"], ["help", "VB"], ["you", "PRP"], ["optimize", "VB"], ["your", "PRP$"], ["nutrition.", "JJ"], ["Steer", "VB"], ["clear", "JJ"], ["of", "IN"], ["", ")"], ["``", "``"], ["natural", "JJ"], ["cures", "NNS"], ["''", "''"], ["", ")"], ["Before", "IN"], ["trying", "VBG"], ["nutritional", "JJ"], ["supplements", "NNS"], ["or", "CC"], ["herbal", "JJ"], ["remedies", "NNS"], [",", ","], ["", ")"], ["be", "VB"], ["sure", "JJ"], ["to", "TO"], ["discuss", "VB"], ["your", "PRP$"], ["plans", "NNS"], ["with", "IN"], ["a", "DT"], ["doctor.", "JJ"], ["Most", "JJS"], ["have", "VBP"], ["not", "RB"], ["been", "VBN"], ["tested", "VBN"], ["in", "IN"], ["clinical", "JJ"], ["studies", "NNS"], [",", ","], ["", ")"], ["and", "CC"], ["some", "DT"], ["may", "MD"], ["actually", "RB"], ["interfere", "VB"], ["with", "IN"], ["your", "PRP$"], ["treatment.", "JJ"], ["Build", "VB"], ["a", "DT"], ["stronger", "JJR"], ["body", "NN"], ["Even", "RB"], ["walking", "VBG"], ["regularly", "RB"], ["is", "VBZ"], ["can", "MD"], ["help", "VB"], ["you", "PRP"], ["minimize", "VB"], ["long-term", "JJ"], ["muscle", "NN"], ["weakness", "NN"], ["caused", "VBN"], ["by", "IN"], ["illness", "NN"], ["or", "CC"], ["de-conditioning.", "NNP"], ["Focus", "NNP"], ["on", "IN"], ["overall", "JJ"], ["health", "NN"], ["Patients", "NNS"], ["may", "MD"], ["be", "VB"], ["cured", "VBN"], ["of", "IN"], ["cancer", "NN"], ["but", "CC"], ["still", "JJ"], ["face", "NN"], ["life-threatening", "JJ"], ["medical", "JJ"], ["problems", "NNS"], ["that", "WDT"], ["are", "VBP"], ["underemphasized", "JJ"], ["during", "IN"], ["cancer", "NN"], ["treatments", "NNS"], [",", ","], ["", ")"], ["such", "JJ"], ["as", "IN"], ["diabetes", "NN"], [",", ","], ["", ")"], ["high", "JJ"], ["blood", "NN"], ["pressure", "NN"], ["and", "CC"], ["heart", "NN"], ["disease.", "JJ"], ["Continue", "VB"], ["to", "TO"], ["monitor", "VB"], ["your", "PRP$"], ["overall", "JJ"], ["health.", "JJ"], ["Put", "NN"], ["the", "DT"], ["fire", "NN"], ["out", "IN"], ["for", "IN"], ["good", "JJ"], ["Smoking", "NNP"], ["impairs", "NNS"], ["healing", "NN"], ["after", "IN"], ["surgery", "NN"], ["and", "CC"], ["radiation", "NN"], ["and", "CC"], ["increases", "NNS"], ["your", "PRP$"], ["risk", "NN"], ["of", "IN"], ["cardiovascular", "JJ"], ["disease", "NN"], ["and", "CC"], ["many", "JJ"], ["types", "NNS"], ["of", "IN"], ["cancers.", "CD"], ["Ask", "VB"], ["your", "PRP$"], ["doctor", "NN"], ["for", "IN"], ["help", "NN"], ["identifying", "VBG"], ["and", "CC"], ["obtaining", "VBG"], ["the", "DT"], ["most", "RBS"], ["appropriate", "JJ"], ["cessation", "NN"], ["aids.", "NNP"], ["Map", "NNP"], ["a", "DT"], ["healthy", "JJ"], ["future", "NN"], ["Once", "RB"], ["youve", "VBP"], ["completed", "VBN"], ["treatment", "NN"], [",", ","], ["", ")"], ["discuss", "VB"], ["appropriate", "JJ"], ["follow-up", "NN"], ["plans", "NNS"], ["with", "IN"], ["your", "PRP$"], ["doctor", "NN"], ["and", "CC"], ["keep", "VB"], ["track", "NN"], ["of", "IN"], ["them", "PRP"], ["yourself.", "CD"], ["Intensified", "JJ"], ["screening", "NN"], ["over", "IN"], ["many", "JJ"], ["years", "NNS"], ["is", "VBZ"], ["frequently", "RB"], ["recommended", "VBN"], ["to", "TO"], ["identify", "VB"], ["and", "CC"], ["treat", "VB"], ["a", "DT"], ["recurrence", "NN"], ["early", "JJ"], ["on.", "CD"], ["Share", "VB"], ["your", "PRP$"], ["feelings", "NNS"], ["Allow", "VB"], ["yourself", "PRP"], ["time", "NN"], ["to", "TO"], ["discuss", "VB"], ["the", "DT"], ["emotional", "JJ"], ["consequences", "NNS"], ["of", "IN"], ["your", "PRP$"], ["illness", "NN"], ["and", "CC"], ["treatment", "NN"], ["with", "IN"], ["family", "NN"], [",", ","], ["", ")"], ["friends", "NNS"], [",", ","], ["", ")"], ["your", "PRP$"], ["doctor", "NN"], ["and", "CC"], [",", ","], ["", ")"], ["if", "IN"], ["necessary", "JJ"], [",", ","], ["", ")"], ["a", "DT"], ["professional", "JJ"], ["therapist.", "JJ"], ["Many", "JJ"], ["patients", "NNS"], ["also", "RB"], ["find", "VBP"], ["antidepressants", "NNS"], ["helpful", "JJ"], ["during", "IN"], ["treatment.", "JJ"], ["Stay", "VB"], ["connected", "VBN"], ["Although", "IN"], ["many", "JJ"], ["newly", "RB"], ["diagnosed", "VBN"], ["patients", "NNS"], ["fear", "VBP"], ["they", "PRP"], ["will", "MD"], ["not", "RB"], ["be", "VB"], ["able", "JJ"], ["to", "TO"], ["keep", "VB"], ["working", "VBG"], ["during", "IN"], ["treatment", "NN"], [",", ","], ["", ")"], ["this", "DT"], ["is", "VBZ"], ["usually", "RB"], ["not", "RB"], ["the", "DT"], ["case.", "CD"], ["Working", "NNP"], [",", ","], ["", ")"], ["even", "RB"], ["at", "IN"], ["a", "DT"], ["reduced", "VBN"], ["schedule", "NN"], [",", ","], ["", ")"], ["helps", "VBZ"], ["you", "PRP"], ["maintain", "VBP"], ["valuable", "JJ"], ["social", "JJ"], ["connections", "NNS"], ["and", "CC"], ["weekly", "JJ"], ["structure", "NN"], [".", "."]], pairs #puts pairs.inspect - #$tagger = nil - #ObjectSpace.garbage_collect + # enable these lines for memory leak testing + $tagger = nil + ObjectSpace.garbage_collect end def test_multiple_docs test/tagger_test.rb @@ -1,60 +1,59 @@ -==29171== -==29171== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 19 from 1) -==29171== malloc/free: in use at exit: 3,284,995 bytes in 22,026 blocks. -==29171== malloc/free: 1,897,908 allocs, 1,875,882 frees, 45,905,687 bytes allocated. -==29171== For counts of detected errors, rerun with: -v -==29171== searching for pointers to 22,026 not-freed blocks. -==29171== checked 3,458,688 bytes. -==29171== -==29171== 16 bytes in 1 blocks are definitely lost in loss record 1 of 14 -==29171== at 0x4005868: malloc (vg_replace_malloc.c:207) -==29171== by 0x8072AF6: ruby_xmalloc (gc.c:139) -==29171== by 0x80A2431: local_append (parse.y:5640) -==29171== by 0x80A2555: local_cnt (parse.y:5667) -==29171== by 0x80A386A: special_local_set (parse.y:6228) -==29171== by 0x80A38F2: rb_backref_set (parse.y:6252) -==29171== by 0x80AB3B9: rb_reg_search (re.c:946) -==29171== by 0x80CD725: rb_str_split_m (string.c:3559) -==29171== by 0x805FB9F: call_cfunc (eval.c:5700) -==29171== by 0x805F109: rb_call0 (eval.c:5856) -==29171== by 0x80606BB: rb_call (eval.c:6103) -==29171== by 0x805948A: rb_eval (eval.c:3479) -==29171== -==29171== -==29171== 26 bytes in 9 blocks are definitely lost in loss record 2 of 14 -==29171== at 0x4005868: malloc (vg_replace_malloc.c:207) -==29171== by 0xB2DE9F: strdup (in /lib/libc-2.7.so) -==29171== by 0x400D11B: tagger_context_add_goodright (tagger.c:105) -==29171== by 0x400DA92: BrillTagger_add_goodright (rbtagger.c:76) -==29171== by 0x805FBD4: call_cfunc (eval.c:5706) -==29171== by 0x805F109: rb_call0 (eval.c:5856) -==29171== by 0x80606BB: rb_call (eval.c:6103) -==29171== by 0x805948A: rb_eval (eval.c:3479) -==29171== by 0x8057D9A: rb_eval (eval.c:3133) -==29171== by 0x805F82F: rb_call0 (eval.c:6007) -==29171== by 0x80606BB: rb_call (eval.c:6103) -==29171== by 0x805948A: rb_eval (eval.c:3479) -==29171== -==29171== -==29171== 1,044 bytes in 33 blocks are possibly lost in loss record 11 of 14 -==29171== at 0x4005868: malloc (vg_replace_malloc.c:207) -==29171== by 0x8072AF6: ruby_xmalloc (gc.c:139) -==29171== by 0x806484F: scope_dup (eval.c:8211) -==29171== by 0x80651D8: proc_alloc (eval.c:8484) -==29171== by 0x806522C: proc_s_new (eval.c:8519) -==29171== by 0x805FB9F: call_cfunc (eval.c:5700) -==29171== by 0x805F109: rb_call0 (eval.c:5856) -==29171== by 0x80606BB: rb_call (eval.c:6103) -==29171== by 0x805948A: rb_eval (eval.c:3479) -==29171== by 0x805A5CA: rb_eval (eval.c:3664) -==29171== by 0x805F72F: rb_call0 (eval.c:5980) -==29171== by 0x80606BB: rb_call (eval.c:6103) -==29171== -==29171== LEAK SUMMARY: -==29171== definitely lost: 42 bytes in 10 blocks. -==29171== possibly lost: 1,044 bytes in 33 blocks. -==29171== still reachable: 3,283,909 bytes in 21,983 blocks. -==29171== suppressed: 0 bytes in 0 blocks. -==29171== Reachable blocks (those to which a pointer was found) are not shown. -==29171== To see them, rerun with: --leak-check=full --show-reachable=yes - +==29196== +==29196== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 19 from 1) +==29196== malloc/free: in use at exit: 3,131,835 bytes in 14,941 blocks. +==29196== malloc/free: 1,897,909 allocs, 1,882,968 frees, 45,905,816 bytes allocated. +==29196== For counts of detected errors, rerun with: -v +==29196== searching for pointers to 14,941 not-freed blocks. +==29196== checked 3,367,580 bytes. +==29196== +==29196== 16 bytes in 1 blocks are definitely lost in loss record 1 of 14 +==29196== at 0x4005868: malloc (vg_replace_malloc.c:207) +==29196== by 0x8072AF6: ruby_xmalloc (gc.c:139) +==29196== by 0x80A2431: local_append (parse.y:5640) +==29196== by 0x80A2555: local_cnt (parse.y:5667) +==29196== by 0x80A386A: special_local_set (parse.y:6228) +==29196== by 0x80A38F2: rb_backref_set (parse.y:6252) +==29196== by 0x80AB3B9: rb_reg_search (re.c:946) +==29196== by 0x80CD725: rb_str_split_m (string.c:3559) +==29196== by 0x805FB9F: call_cfunc (eval.c:5700) +==29196== by 0x805F109: rb_call0 (eval.c:5856) +==29196== by 0x80606BB: rb_call (eval.c:6103) +==29196== by 0x805948A: rb_eval (eval.c:3479) +==29196== +==29196== +==29196== 26 bytes in 9 blocks are definitely lost in loss record 2 of 14 +==29196== at 0x4005868: malloc (vg_replace_malloc.c:207) +==29196== by 0xB2DE9F: strdup (in /lib/libc-2.7.so) +==29196== by 0x400D11B: tagger_context_add_goodright (tagger.c:105) +==29196== by 0x400DA92: BrillTagger_add_goodright (rbtagger.c:76) +==29196== by 0x805FBD4: call_cfunc (eval.c:5706) +==29196== by 0x805F109: rb_call0 (eval.c:5856) +==29196== by 0x80606BB: rb_call (eval.c:6103) +==29196== by 0x805948A: rb_eval (eval.c:3479) +==29196== by 0x8057D9A: rb_eval (eval.c:3133) +==29196== by 0x805F82F: rb_call0 (eval.c:6007) +==29196== by 0x80606BB: rb_call (eval.c:6103) +==29196== by 0x805948A: rb_eval (eval.c:3479) +==29196== +==29196== +==29196== 736 bytes in 24 blocks are possibly lost in loss record 11 of 14 +==29196== at 0x4005868: malloc (vg_replace_malloc.c:207) +==29196== by 0x8072AF6: ruby_xmalloc (gc.c:139) +==29196== by 0x806484F: scope_dup (eval.c:8211) +==29196== by 0x80651D8: proc_alloc (eval.c:8484) +==29196== by 0x806522C: proc_s_new (eval.c:8519) +==29196== by 0x805FB9F: call_cfunc (eval.c:5700) +==29196== by 0x805F109: rb_call0 (eval.c:5856) +==29196== by 0x80606BB: rb_call (eval.c:6103) +==29196== by 0x805948A: rb_eval (eval.c:3479) +==29196== by 0x805A5CA: rb_eval (eval.c:3664) +==29196== by 0x805F72F: rb_call0 (eval.c:5980) +==29196== by 0x80606BB: rb_call (eval.c:6103) +==29196== +==29196== LEAK SUMMARY: +==29196== definitely lost: 42 bytes in 10 blocks. +==29196== possibly lost: 736 bytes in 24 blocks. +==29196== still reachable: 3,131,057 bytes in 14,907 blocks. +==29196== suppressed: 0 bytes in 0 blocks. +==29196== Reachable blocks (those to which a pointer was found) are not shown. +==29196== To see them, rerun with: --leak-check=full --show-reachable=yes valgrind.leaks 6bfcd04d964c1a1e35f24e9754ceddc197151a3a Todd Fisher taf2@bigbox.localdomain http://github.com/taf2/rb-brill-tagger/commit/56ff1857d0e91dc1dde94249b26c7822c6e63d26 56ff1857d0e91dc1dde94249b26c7822c6e63d26 2008-05-14T06:23:09-07:00 2008-05-14T06:23:09-07:00 update for leak checking 5c1eaeeaffeddeebeb143f83a246f23332763d12 Todd Fisher taf2@bigbox.localdomain