Fix bug in strndup fix for windows. Move all includes out of headers and into code for strndup.h and move it to be the last include.

src/Makefile.am

@@ -12,7 +12,7 @@ DEFAULT_INCLUDES = -I.. -I/usr/local/include
 CFLAGS = -D UTF8PROC_EXPORTS -D LIBPOSTAL_EXPORTS
 
 lib_LTLIBRARIES = libpostal.la
-libpostal_la_SOURCES = strndup.c libpostal.c address_dictionary.c transliterate.c tokens.c trie.c trie_search.c trie_utils.c string_utils.c file_utils.c utf8proc/utf8proc.c cmp/cmp.c normalize.c numex.c features.c unicode_scripts.c address_parser.c address_parser_io.c averaged_perceptron.c crf.c crf_context.c sparse_matrix.c averaged_perceptron_tagger.c graph.c graph_builder.c language_classifier.c language_features.c logistic_regression.c logistic.c minibatch.c float_utils.c ngrams.c
+libpostal_la_SOURCES = libpostal.c address_dictionary.c transliterate.c tokens.c trie.c trie_search.c trie_utils.c string_utils.c file_utils.c utf8proc/utf8proc.c cmp/cmp.c normalize.c numex.c features.c unicode_scripts.c address_parser.c address_parser_io.c averaged_perceptron.c crf.c crf_context.c sparse_matrix.c averaged_perceptron_tagger.c graph.c graph_builder.c language_classifier.c language_features.c logistic_regression.c logistic.c minibatch.c float_utils.c ngrams.c
 libpostal_la_LIBADD = libscanner.la $(CBLAS_LIBS)
 libpostal_la_CFLAGS = $(CFLAGS_O2)
 libpostal_la_LDFLAGS = -version-info @LIBPOSTAL_SO_VERSION@ -no-undefined
@@ -38,27 +38,27 @@ address_parser_SOURCES = address_parser_cli.c json_encode.c linenoise/linenoise.
 address_parser_LDADD = libscanner.la $(CBLAS_LIBS)
 address_parser_CFLAGS = $(CFLAGS_O3)
 
-build_address_dictionary_SOURCES = strndup.c address_dictionary_builder.c address_dictionary.c file_utils.c string_utils.c trie.c trie_search.c utf8proc/utf8proc.c
+build_address_dictionary_SOURCES = address_dictionary_builder.c address_dictionary.c file_utils.c string_utils.c trie.c trie_search.c utf8proc/utf8proc.c
 build_address_dictionary_CFLAGS = $(CFLAGS_O3)
-build_numex_table_SOURCES = strndup.c numex_table_builder.c numex.c file_utils.c string_utils.c tokens.c trie.c trie_search.c utf8proc/utf8proc.c
+build_numex_table_SOURCES = numex_table_builder.c numex.c file_utils.c string_utils.c tokens.c trie.c trie_search.c utf8proc/utf8proc.c
 build_numex_table_CFLAGS = $(CFLAGS_O3)
-build_trans_table_SOURCES = strndup.c transliteration_table_builder.c transliterate.c trie.c trie_search.c file_utils.c string_utils.c utf8proc/utf8proc.c
+build_trans_table_SOURCES = transliteration_table_builder.c transliterate.c trie.c trie_search.c file_utils.c string_utils.c utf8proc/utf8proc.c
 build_trans_table_CFLAGS = $(CFLAGS_O3)
-address_parser_train_SOURCES = strndup.c address_parser_train.c address_parser.c address_parser_io.c averaged_perceptron.c crf.c crf_context.c sparse_matrix.c graph.c graph_builder.c float_utils.c averaged_perceptron_trainer.c crf_trainer.c crf_trainer_averaged_perceptron.c averaged_perceptron_tagger.c address_dictionary.c normalize.c numex.c features.c unicode_scripts.c transliterate.c trie.c trie_search.c trie_utils.c string_utils.c tokens.c file_utils.c shuffle.c utf8proc/utf8proc.c ngrams.c
+address_parser_train_SOURCES = address_parser_train.c address_parser.c address_parser_io.c averaged_perceptron.c crf.c crf_context.c sparse_matrix.c graph.c graph_builder.c float_utils.c averaged_perceptron_trainer.c crf_trainer.c crf_trainer_averaged_perceptron.c averaged_perceptron_tagger.c address_dictionary.c normalize.c numex.c features.c unicode_scripts.c transliterate.c trie.c trie_search.c trie_utils.c string_utils.c tokens.c file_utils.c shuffle.c utf8proc/utf8proc.c ngrams.c
 address_parser_train_LDADD = libscanner.la $(CBLAS_LIBS)
 address_parser_train_CFLAGS = $(CFLAGS_O3)
 
-address_parser_test_SOURCES = strndup.c address_parser_test.c address_parser.c address_parser_io.c averaged_perceptron.c crf.c crf_context.c sparse_matrix.c graph.c graph_builder.c float_utils.c averaged_perceptron_tagger.c address_dictionary.c normalize.c numex.c features.c unicode_scripts.c transliterate.c trie.c trie_search.c trie_utils.c string_utils.c tokens.c file_utils.c utf8proc/utf8proc.c ngrams.c
+address_parser_test_SOURCES = address_parser_test.c address_parser.c address_parser_io.c averaged_perceptron.c crf.c crf_context.c sparse_matrix.c graph.c graph_builder.c float_utils.c averaged_perceptron_tagger.c address_dictionary.c normalize.c numex.c features.c unicode_scripts.c transliterate.c trie.c trie_search.c trie_utils.c string_utils.c tokens.c file_utils.c utf8proc/utf8proc.c ngrams.c
 address_parser_test_LDADD = libscanner.la  $(CBLAS_LIBS)
 address_parser_test_CFLAGS = $(CFLAGS_O3)
 
-language_classifier_train_SOURCES = strndup.c language_classifier_train.c language_classifier.c language_features.c language_classifier_io.c logistic_regression_trainer.c logistic_regression.c logistic.c sparse_matrix.c sparse_matrix_utils.c features.c minibatch.c float_utils.c stochastic_gradient_descent.c ftrl.c regularization.c cartesian_product.c normalize.c numex.c transliterate.c trie.c trie_search.c trie_utils.c address_dictionary.c string_utils.c file_utils.c utf8proc/utf8proc.c unicode_scripts.c shuffle.c
+language_classifier_train_SOURCES = language_classifier_train.c language_classifier.c language_features.c language_classifier_io.c logistic_regression_trainer.c logistic_regression.c logistic.c sparse_matrix.c sparse_matrix_utils.c features.c minibatch.c float_utils.c stochastic_gradient_descent.c ftrl.c regularization.c cartesian_product.c normalize.c numex.c transliterate.c trie.c trie_search.c trie_utils.c address_dictionary.c string_utils.c file_utils.c utf8proc/utf8proc.c unicode_scripts.c shuffle.c
 language_classifier_train_LDADD = libscanner.la $(CBLAS_LIBS)
 language_classifier_train_CFLAGS = $(CFLAGS_O3)
-language_classifier_SOURCES = strndup.c language_classifier_cli.c language_classifier.c language_features.c logistic_regression.c logistic.c sparse_matrix.c features.c minibatch.c float_utils.c normalize.c numex.c transliterate.c trie.c trie_search.c trie_utils.c address_dictionary.c string_utils.c file_utils.c utf8proc/utf8proc.c unicode_scripts.c
+language_classifier_SOURCES = language_classifier_cli.c language_classifier.c language_features.c logistic_regression.c logistic.c sparse_matrix.c features.c minibatch.c float_utils.c normalize.c numex.c transliterate.c trie.c trie_search.c trie_utils.c address_dictionary.c string_utils.c file_utils.c utf8proc/utf8proc.c unicode_scripts.c
 language_classifier_LDADD = libscanner.la $(CBLAS_LIBS)
 language_classifier_CFLAGS = $(CFLAGS_O3)
-language_classifier_test_SOURCES = strndup.c language_classifier_test.c language_classifier.c language_classifier_io.c language_features.c logistic_regression.c logistic.c sparse_matrix.c features.c minibatch.c float_utils.c normalize.c numex.c transliterate.c trie.c trie_search.c trie_utils.c address_dictionary.c string_utils.c file_utils.c utf8proc/utf8proc.c unicode_scripts.c
+language_classifier_test_SOURCES = language_classifier_test.c language_classifier.c language_classifier_io.c language_features.c logistic_regression.c logistic.c sparse_matrix.c features.c minibatch.c float_utils.c normalize.c numex.c transliterate.c trie.c trie_search.c trie_utils.c address_dictionary.c string_utils.c file_utils.c utf8proc/utf8proc.c unicode_scripts.c
 language_classifier_test_LDADD = libscanner.la $(CBLAS_LIBS)
 language_classifier_test_CFLAGS = $(CFLAGS_O3)
 

src/normalize.c

@@ -1,4 +1,5 @@
 #include "normalize.h"
+#include "strndup.h"
 
 #define FULL_STOP_CODEPOINT 0x002e
 #define APOSTROPHE_CODEPOINT 0x0027

src/normalize.h

@@ -38,7 +38,6 @@ As well as normalizations for individual string tokens:
 #include "trie.h"
 #include "tokens.h"
 #include "vector.h"
-#include "strndup.h"
 
 #define NORMALIZE_STRING_LATIN_ASCII 1 << 0
 #define NORMALIZE_STRING_TRANSLITERATE 1 << 1

src/string_utils.c

@@ -1,6 +1,7 @@
 #include <stdio.h>
 #include "log/log.h"
 #include "string_utils.h"
+#include "strndup.h"
 
 #define INVALID_INDEX(i, n) ((i) < 0 || (i) >= (n))
 

src/string_utils.h

@@ -16,7 +16,6 @@ Utilities for manipulating strings in C.
 #include "collections.h"
 #include "utf8proc/utf8proc.h"
 #include "vector.h"
-#include "strndup.h"
 #include "export.h"
 
 #define MAX_UTF8_CHAR_SIZE 4

src/strndup.c

@@ -13,4 +13,4 @@ char *strndup(const char *s, size_t n)
     return new;
 }
 
-#endif /* HAVE_STRNDUP */
+#endif /* HAVE_STRNDUP */

src/strndup.h

@@ -1,6 +1,11 @@
+#ifndef STRNDUP_H
+#define STRNDUP_H
+
+#include <config.h>
+
 #ifndef HAVE_STRNDUP
-#define HAVE_STRNDUP
 
 char *strndup(const char *s, size_t n);
 
-#endif /* HAVE_STRNDUP */
+#endif /* HAVE_STRNDUP */
+#endif /* STRNDUP_H */

src/tokens.c

@@ -1,4 +1,5 @@
 #include "tokens.h"
+#include "strndup.h"
 
 
 tokenized_string_t *tokenized_string_new(void) {

src/tokens.h

@@ -1,7 +1,7 @@
 #ifndef TOKENS_H
 #define TOKENS_H
 
- 
+
 
 #include <stdlib.h>
 #include <stdint.h>
@@ -11,7 +11,6 @@
 #include "string_utils.h"
 #include "token_types.h"
 #include "vector.h"
-#include "strndup.h"
 
 typedef struct token {
     size_t offset;
@@ -36,6 +35,6 @@ char *tokenized_string_get_token(tokenized_string_t *self, uint32_t index);
 void tokenized_string_destroy(tokenized_string_t *self);
 
 
- 
+
 
 #endif

src/transliterate.c

@@ -3,6 +3,7 @@
 #include "file_utils.h"
 
 #include "log/log.h"
+#include "strndup.h"
 
 #define TRANSLITERATION_TABLE_SIGNATURE 0xAAAAAAAA
 

src/transliterate.h

@@ -12,7 +12,6 @@
 #include "trie.h"
 #include "trie_search.h"
 #include "unicode_scripts.h"
-#include "strndup.h"
 #include "export.h"
 
 #define LATIN_ASCII "latin-ascii"