Improve deinflector

Author: btrkeks

Makefile

@@ -9,22 +9,22 @@ SDIR=src
 LDIR=lib
 
 CPPFLAGS += -D_POSIX_C_SOURCE=200809L -DNOTIFICATIONS -I$(IDIR)
-LDLIBS += -ffunction-sections -fdata-sections -Wl,--gc-sections \
-       	  -lcurl -lmecab -pthread $(shell pkg-config --libs gtk+-3.0) $(shell pkg-config --libs libnotify) -llmdb
+LDLIBS +=-lcurl -lmecab -pthread $(shell pkg-config --libs gtk+-3.0) $(shell pkg-config --libs libnotify) -llmdb
+# LDLIBS+=-ffunction-sections -fdata-sections -Wl,--gc-sections
 
 O_HAVEX11 := 1  # X11 integration
 ifeq ($(strip $(O_HAVEX11)),1)
 	CPPFLAGS += -DHAVEX11
 	LDLIBS += -lXfixes -lX11
 endif
 
-CFLAGS := -Werror $(shell pkg-config --cflags gtk+-3.0) $(shell pkg-config --cflags libnotify) $(CFLAGS)
+CFLAGS := -Werror $(shell pkg-config --cflags gtk+-3.0) $(shell pkg-config --cflags libnotify)
 DEBUG_CFLAGS=-DDEBUG \
 	     -Wall -Wextra -Wpedantic -Wstrict-prototypes -Wdouble-promotion -Wshadow \
 	     -Wno-unused-parameter -Wno-sign-conversion -Wno-unused-function -Wpointer-arith \
 	     -Wmissing-prototypes -Wstrict-prototypes -Wstrict-overflow -Wcast-align \
 	     -fsanitize=leak,address,undefined -fsanitize-undefined-trap-on-error -fstack-protector-strong \
-	     -Og -ggdb
+	     -O0 -ggdb
 RELEASE_CFLAGS=-O3 -flto -march=native -Wmaybe-uninitialized
 
 FILES=dictpopup.c util.c platformdep.c deinflector.c settings.c db.c ankiconnectc.c database.c jppron.c pdjson.c
@@ -33,12 +33,11 @@ SRC=$(addprefix $(SDIR)/,$(FILES))
 SRC_H=$(addprefix $(IDIR)/,$(FILES_H))
 
 
-CFLAGS_CREATE=-I$(IDIR) -isystem$(LIBDIR)/lmdb/libraries/liblmdb -D_POSIX_C_SOURCE=200809L $(shell pkg-config --cflags glib-2.0)
-LDLIBS_CREATE=-ffunction-sections -fdata-sections -Wl,--gc-sections \
-	      -lzip $(shell pkg-config --libs glib-2.0) -llmdb
+CFLAGS_CREATE=-I$(IDIR) -D_POSIX_C_SOURCE=200809L $(shell pkg-config --cflags glib-2.0)
+LDLIBS_CREATE=-lzip $(shell pkg-config --libs glib-2.0) -llmdb
 
-FILES_CREATE=db.c pdjson.c util.c settings.c
-FILES_H_CREATE=db.h pdjson.h util.h buf.h settings.h
+FILES_CREATE=db.c pdjson.c util.c
+FILES_H_CREATE=db.h pdjson.h util.h buf.h
 
 SRC_CREATE=$(addprefix $(SDIR)/,$(FILES_CREATE))  $(LMDB_FILES)
 SRC_H_CREATE=$(addprefix $(IDIR)/,$(FILES_H_CREATE))
@@ -57,6 +56,11 @@ dictpopup: $(SRC) $(SRC_H)
 dictpopup-create: $(SRC_CREATE) $(SRC_H_CREATE)
 	$(CC) $(CFLAGS_CREATE) -o $@ $(SDIR)/dictpopup_create.c $(SRC_CREATE) $(LDLIBS_CREATE)
 
+cli: $(SRC) $(SRC_H)
+	$(CC) $(CFLAGS) $(DEBUG_CFLAGS) $(CPPFLAGS) -o $@ $(SDIR)/cli.c $(SRC) $(LDLIBS)
+deinflector: $(SRC) $(SRC_H)
+	$(CC) $(CFLAGS) $(DEBUG_CFLAGS) -I$(IDIR) -DDEINFLECTOR_MAIN -o $@ $(SDIR)/deinflector.c $(SDIR)/util.c $(LDLIBS)
+
 release:
 	version=$$(git describe); prefix=dictpopup-$${version#v}; \
 	  git archive --prefix=$$prefix/ HEAD | gzip -9 >$$prefix.tar.gz
@@ -86,7 +90,7 @@ uninstall:
 	      $(CONFIG_DIR)/config.ini
 
 clean:
-	rm -f dictpopup dictpopup-create tests
+	rm -f dictpopup dictpopup-create tests cli deinflector
 
 tests: $(SRC) $(SRC_H)
 	$(CC) $(CFLAGS) $(DEBUG_CFLAGS) $(CPPFLAGS) -o $@ $(SDIR)/tests.c $(SRC) $(LDLIBS)

TODO

@@ -1,16 +1,13 @@
 ## dictpopup
-High priority
-- Refactor
-  - Allow for an easier replacement of the frontend
-    - Expose API to frontend like init() and end()?
-  -> Write a "Yomichan search"-like window application reusing the same backend
-- Create config file in ~/.config if not existent
+- Fix formating for JMDict and Jitendex
+- Write a "Yomichan search"-like window application
+- Include a small default dictionary (.mdb file)
+- Add ながら deinflection
 
 Mid priority
 - Implement the reading label as a text view to allow for editing -> initiate new search on Enter press
 - Group terms by reading, allow for switching between different readings
 - Implement a small window to manually enter a search string instead of selecting
-- Expand deinflection list
 - Kanjify the word in the copied sentence according to the dictionary entry
 - Remove trailing spaces from user settings
 - Write a gui config tool
@@ -47,7 +44,6 @@ Low priority
 - Skip certain lines on definition add (without selection), e.g.:
   - しず か （しづ—） [1]【静か・▽閑か】
   -（形動）〔文〕ナリ
-- Add unit tests and do fuzzy testing
 - Put length of longest entry into the database under LENGTH_OF_LONGEST_ENTRY. Then limit the lookup by that length
   - Only important for an auto scan implementation though
 
@@ -59,21 +55,15 @@ Low priority
 ### Misc
 - If there is a dictionary entry for 恭しく even though 恭しい is present in the collection, then a red dot will be shown
 ### Deinflection
-- 与ろうとする missing
 - やった gets deinflected to やつ
   - This can be fixed by reading in the 4th string of a Yomichan term-bank entry, which is:
     "String of space-separated rule identifiers for the definition which is used to validate deinflection.
     An empty string should be used for words which aren't inflected."
-- たいそう gets deinflected to たいい
-- 懐から becomes 懐く
-- 白けた gets deinflected to 白い
-- 目した only gets deinflected to 目す
-- はたきながら (should become はたく)
-- 挿げ(変わる)
 - に looks broken
+
+Mecab problems:
 - 大正門 pulls 大誓文
-- Can't definlect 眠ろうとする, 粘ろうとする
-- 失われ、
+- 家名だ
 
 freq:
 - いじらしい not present in all

src/cli.c

@@ -0,0 +1,12 @@
+#include "dictpopup.h"
+#include "util.h"
+
+int main(int argc, char **argv) {
+    dictpopup_t d = dictpopup_init(argc, argv);
+    dictentry *dict = create_dictionary(&d);
+
+    for (size i = 0; i < dictlen(dict); i++) {
+        dictentry_print(dict[i]);
+        putchar('\n');
+    }
+}

src/db.c

@@ -27,6 +27,7 @@ database_t db_open(char *dbpath, bool readonly) {
         unsigned int mapsize = 2097152000; // 2Gb
         C(mdb_env_set_mapsize(db.env, mapsize));
 
+        dbg("Opening db in path: %s", dbpath);
         C(mdb_env_open(db.env, dbpath, 0, 0664));
         C(mdb_txn_begin(db.env, NULL, 0, &db.txn));
 
@@ -154,6 +155,7 @@ static s8 getdata(database_t *db, u32 id) {
 }
 
 void db_get_dictents(database_t *db, s8 headword, dictentry *dict[static 1]) {
+    dbg("Looking up: %.*s", (int)headword.len, (char *)headword.s);
 
     size_t n_ids = 0;
     u32 *ids = get_ids(db, headword, &n_ids);