| summaryrefslogtreecommitdiff | 
diff options
| author | Nathanael Sensfelder <SpamShield0@MultiAgentSystems.org> | 2017-01-18 19:09:16 +0100 | 
|---|---|---|
| committer | Nathanael Sensfelder <SpamShield0@MultiAgentSystems.org> | 2017-01-18 19:09:16 +0100 | 
| commit | 0d49fb74eadcf933f696420cd182077927680d26 (patch) | |
| tree | 9220d260ce878f369138da12dae0300cf9ade5c9 /src/core | |
| parent | 24afb3e60bafd98e6a83dcb41ee6a7f7d41e76bc (diff) | |
Done with 'core', starting to work on 'knowledge'.
Diffstat (limited to 'src/core')
| -rw-r--r-- | src/core/CMakeLists.txt | 9 | ||||
| -rw-r--r-- | src/core/assimilate.c | 281 | ||||
| -rw-r--r-- | src/core/char.c | 18 | ||||
| -rw-r--r-- | src/core/char.h | 35 | ||||
| -rw-r--r-- | src/core/char_types.h | 6 | ||||
| -rw-r--r-- | src/core/index.c | 61 | ||||
| -rw-r--r-- | src/core/index.h | 11 | ||||
| -rw-r--r-- | src/core/index_types.h | 2 | ||||
| -rw-r--r-- | src/core/knowledge.c | 15 | ||||
| -rw-r--r-- | src/core/knowledge.h | 101 | ||||
| -rw-r--r-- | src/core/knowledge_finalize.c | 121 | ||||
| -rw-r--r-- | src/core/knowledge_search.c | 335 | ||||
| -rw-r--r-- | src/core/knowledge_types.h | 37 | ||||
| -rw-r--r-- | src/core/main.c | 436 | ||||
| -rw-r--r-- | src/core/sequence.c | 84 | ||||
| -rw-r--r-- | src/core/sequence.h | 20 | ||||
| -rw-r--r-- | src/core/sequence_creation.c | 9 | ||||
| -rw-r--r-- | src/core/sequence_from_string.c | 315 | ||||
| -rw-r--r-- | src/core/sequence_types.h | 3 | 
19 files changed, 526 insertions, 1373 deletions
| diff --git a/src/core/CMakeLists.txt b/src/core/CMakeLists.txt index fe28080..1e1daa8 100644 --- a/src/core/CMakeLists.txt +++ b/src/core/CMakeLists.txt @@ -1,13 +1,10 @@  set(     SRC_FILES ${SRC_FILES}     ${CMAKE_CURRENT_SOURCE_DIR}/char.c -   ${CMAKE_CURRENT_SOURCE_DIR}/main.c -   ${CMAKE_CURRENT_SOURCE_DIR}/knowledge.c -   ${CMAKE_CURRENT_SOURCE_DIR}/knowledge_search.c -   ${CMAKE_CURRENT_SOURCE_DIR}/knowledge_finalize.c -   ${CMAKE_CURRENT_SOURCE_DIR}/assimilate.c -   ${CMAKE_CURRENT_SOURCE_DIR}/sequence_creation.c +   ${CMAKE_CURRENT_SOURCE_DIR}/index.c     ${CMAKE_CURRENT_SOURCE_DIR}/sequence.c +   ${CMAKE_CURRENT_SOURCE_DIR}/sequence_creation.c +   ${CMAKE_CURRENT_SOURCE_DIR}/sequence_from_string.c  )  set(SRC_FILES ${SRC_FILES} PARENT_SCOPE) diff --git a/src/core/assimilate.c b/src/core/assimilate.c deleted file mode 100644 index 7f03e1b..0000000 --- a/src/core/assimilate.c +++ /dev/null @@ -1,281 +0,0 @@ -#include <stdlib.h> -#include <string.h> - -#include "../io/error.h" - -#include "knowledge.h" - -/** Functions to assimilate sentences using a ZoO_knowledge structure *********/ - - -static int add_sequence -( -   ZoO_index links_count [const], -   struct ZoO_knowledge_link * links [const], -   ZoO_index const sequence [const restrict static ZoO_MARKOV_ORDER], -   ZoO_index const target_i, -   ZoO_index const offset -) -{ -   ZoO_index link_index, i; -   struct ZoO_knowledge_link * link; -   ZoO_index * new_p; - -   if -   ( -      ZoO_knowledge_get_link -      ( -         links_count, -         links, -         (sequence + offset), -         &link_index -      ) < 0 -   ) -   { -      return -1; -   } - -   link = (*links + link_index); -   link->occurrences += 1; - -   for (i = 0; i < link->targets_count; ++i) -   { -      if (link->targets[i] == sequence[target_i]) -      { -         link->targets_occurrences[i] += 1; - -         return 0; -      } -   } - -   link->targets_count += 1; - -   new_p = -      (ZoO_index *) realloc -      ( -         (void *) link->targets, -         (sizeof(ZoO_index) * link->targets_count) -      ); - -   if (new_p == (ZoO_index *) NULL) -   { -      link->targets_count -= 1; - -      /* TODO: err. */ -      return -1; -   } - -   link->targets = new_p; -   link->targets[link->targets_count - 1] = sequence[target_i]; - -   new_p = -      (ZoO_index *) realloc -      ( -         (void *) link->targets_occurrences, -         (sizeof(ZoO_index) * link->targets_count) -      ); - -   if (new_p == (ZoO_index *) NULL) -   { -      link->targets_count -= 1; - -      /* TODO: err. */ -      return -1; -   } - -   link->targets_occurrences = new_p; -   link->targets_occurrences[link->targets_count - 1] = 1; - -   return 0; -} - -static int add_word_occurrence -( -   struct ZoO_knowledge k [const restrict static 1], -   ZoO_index const sequence [const static ((ZoO_MARKOV_ORDER * 2) + 1)] -) -{ -   ZoO_index w; -   int error; - -   w = sequence[ZoO_MARKOV_ORDER]; - -   error = -      add_sequence -      ( -         &(k->words[w].forward_links_count), -         &(k->words[w].forward_links), -         sequence + (ZoO_MARKOV_ORDER + 1), -         (ZoO_MARKOV_ORDER - 1), -         0 -      ); - -   error = -      ( -         add_sequence -         ( -            &(k->words[w].backward_links_count), -            &(k->words[w].backward_links), -            sequence, -            0, -            1 -         ) -         | error -      ); - -   return error; -} - -static int should_assimilate -( -   struct ZoO_strings string [const restrict static 1], -   ZoO_index const aliases_count, -   const char * restrict aliases [const restrict static aliases_count] -) -{ -   ZoO_index i; - -   /* Don't assimilate empty strings. */ -   if (string->words_count == 0) -   { -      return 0; -   } - -   /* Don't assimilate things that start with our name. */ -   for (i = 0; i < aliases_count; ++i) -   { -      if (ZoO_IS_PREFIX(aliases[i], string->words[0])) -      { -         return 0; -      } -   } - -   return 1; -} - -static int init_sequence -( -   struct ZoO_knowledge k [const static 1], -   struct ZoO_strings string [const restrict static 1], -   ZoO_index sequence [const restrict static ((ZoO_MARKOV_ORDER * 2) + 1)] -) -{ -   ZoO_index i; - -   /* We are going to link this sequence to ZoO_WORD_START_OF_LINE */ -   sequence[ZoO_MARKOV_ORDER] = ZoO_WORD_START_OF_LINE; - -   for (i = 1; i <= ZoO_MARKOV_ORDER; ++i) -   { -      sequence[ZoO_MARKOV_ORDER - i] = ZoO_WORD_START_OF_LINE; - -      if (i <= string->words_count) -      { -         if -         ( -            ZoO_knowledge_learn -            ( -               k, -               string->words[i - 1], -               (sequence + (ZoO_MARKOV_ORDER + i)) -            ) < 0 -         ) -         { -            return -1; -         } -      } -      else -      { -         sequence[ZoO_MARKOV_ORDER + i] = ZoO_WORD_END_OF_LINE; -      } -   } - -   return 0; -} - -int ZoO_knowledge_assimilate -( -   struct ZoO_knowledge k [const static 1], -   struct ZoO_strings string [const restrict static 1], -   ZoO_index const aliases_count, -   const char * restrict aliases [const restrict static aliases_count] -) -{ -   int error; -   ZoO_index sequence[(ZoO_MARKOV_ORDER * 2) + 1]; -   ZoO_index next_word, new_word, new_word_id; - -   if (!should_assimilate(string, aliases_count, aliases)) -   { -      return 0; -   } - -   if (init_sequence(k, string, sequence) < 0) -   { -      return -1; -   } - -   if (add_word_occurrence(k, sequence) < 0) -   { -      error = -1; - -      /* There's a pun... */ -      ZoO_S_WARNING("Could not add a link between words."); - -      return -1; -   } - -   error = 0; - -   next_word = 0; -   new_word = ZoO_MARKOV_ORDER; - -   while (next_word <= (string->words_count + ZoO_MARKOV_ORDER)) -   { -      if (new_word < string->words_count) -      { -         /* prevents words [restrict], k [restrict] */ -         if (ZoO_knowledge_learn(k, string->words[new_word], &new_word_id) < 0) -         { -            return -1; -         } -      } -      else -      { -         new_word_id = ZoO_WORD_END_OF_LINE; -      } - -      memmove -      ( -         (void *) sequence, -         (const void *) (sequence + 1), -         /* Accepts 0. */ -         (sizeof(ZoO_index) * (ZoO_MARKOV_ORDER * 2)) -      ); - -      sequence[ZoO_MARKOV_ORDER * 2] = new_word_id; - -      if (add_word_occurrence(k, sequence) < 0) -      { -         error = -1; - -         /* There's a pun... */ -         ZoO_S_WARNING("Could not add a link between words."); - -         return -1; -      } - -      /* -       * Safe: -       *  - next_word < words_count -       *  - words_count =< ZoO_INDEX_MAX -       *  ---- -       *  next_word < ZoO_INDEX_MAX -       */ -      next_word += 1; -      new_word += 1; -   } - -   return error; -} - diff --git a/src/core/char.c b/src/core/char.c index 39ca72e..9297643 100644 --- a/src/core/char.c +++ b/src/core/char.c @@ -2,6 +2,18 @@  #include "char.h" +/* See: "char.c" */ +ZoO_char ZoO_char_to_lowercase (const ZoO_char c) +{ +   if ((c >= 'A') && (c <= 'Z')) +   { +      return 'z' - ('Z' - c); +   } + +   return c; +} + +/* See: "char.c" */  int ZoO_char_is_banned (const ZoO_char c)  {     switch (c) @@ -21,6 +33,7 @@ int ZoO_char_is_banned (const ZoO_char c)     }  } +/* See: "char.c" */  int ZoO_char_is_punctuation (const ZoO_char c)  {     switch (c) @@ -38,11 +51,14 @@ int ZoO_char_is_punctuation (const ZoO_char c)     }  } +/* See: "char.c" */  int ZoO_word_cmp  (     const ZoO_char word_a [const static 1], +   const size_t word_a_size,     const ZoO_char word_b [const static 1]  )  { -   return strcmp((const char *) word_a, (const char *) word_b); +   return strncmp((const char *) word_a, (const char *) word_b, word_a_size);  } + diff --git a/src/core/char.h b/src/core/char.h index 772a3a2..2b4a355 100644 --- a/src/core/char.h +++ b/src/core/char.h @@ -3,19 +3,42 @@  #include "char_types.h" -enum ZoO_word_property ZoO_get_word_property -( -   const ZoO_char word [const restrict], -   size_t word_size -); - +/* Compares two words. {word_a} does not have to be null terminated. */ +/*@ + @ requires null_terminated_string(word_b); + @ requires ((length(word_a) * sizeof(ZoO_char)) == word_a_size); + @ ensures ((\result == 1) || (\result == 0) || (\result == -1)); + @*/  int ZoO_word_cmp  (     const ZoO_char word_a [const static 1], +   const size_t word_a_size,     const ZoO_char word_b [const static 1]  ); +/* + * Returns the lowercase equivalent of ZoO_char that are included in ['A','Z']. + * Other ZoO_char are returned untouched. + */ +ZoO_char ZoO_char_to_lowercase (const ZoO_char c); + +/* + * Returns '1' iff {c} should be considered as an punctuation character, '0' + * otherwise. + */ +/*@ + @ ensures ((\result == 1) || (\result == 0)); + @*/  int ZoO_char_is_punctuation (const ZoO_char c); + +/* + * Returns '1' iff containing {c} means the word should not be learned. '0' + * otherwise. + */ +/*@ + @ ensures ((\result == 1) || (\result == 0)); + @*/  int ZoO_word_char_is_banned (const ZoO_char c);  #endif + diff --git a/src/core/char_types.h b/src/core/char_types.h index 67b5294..a2a736c 100644 --- a/src/core/char_types.h +++ b/src/core/char_types.h @@ -1,12 +1,16 @@  #ifndef _ZoO_CORE_CHAR_TYPES_H_  #define _ZoO_CORE_CHAR_TYPES_H_ - +/* + * FIXME: Does not belong here. + */ +/*  enum ZoO_word_property  {     ZoO_WORD_NO_PROPERTY,     ZoO_WORD_HAS_NO_LEFT_SEPARATOR,     ZoO_WORD_HAS_NO_RIGHT_SEPARATOR  }; +*/  /* ZoO_char = UTF-8 char */  typedef char ZoO_char; diff --git a/src/core/index.c b/src/core/index.c new file mode 100644 index 0000000..375e0ad --- /dev/null +++ b/src/core/index.c @@ -0,0 +1,61 @@ +#include <limits.h> +#include <stdlib.h> + +#include "index.h" + +#if (RAND_MAX < UCHAR_MAX) +   #error "RAND_MAX < UCHAR_MAX, unable to generate random numbers." +#endif + +#if (RAND_MAX == 0) +   #error "RAND_MAX is included in [0, 0]. What are you even doing?" +#endif + +/* + * Returns a random unsigned char. + */ +static unsigned char random_uchar (void) +{ +   return +   (unsigned char) +   ( +      /* FIXME: Do floats allow enough precision for this? */ +      ( +         ((float) rand()) +         / ((float) RAND_MAX) +      ) +      * ((float) UCHAR_MAX) +   ); +} + +/* See: "index.h" */ +ZoO_index ZoO_index_random (void) +{ +   ZoO_index i; +   ZoO_index result; +   unsigned char * result_bytes; + +   result_bytes = (unsigned char *) &result; + +   for (i = 0; i < sizeof(ZoO_index); ++i) +   { +      result_bytes[i] = random_uchar(); +   } + +   return result; +} + +/* See: "index.h" */ +ZoO_index ZoO_index_random_up_to (const ZoO_index max) +{ +   return +   (ZoO_index) +   ( +      /* FIXME: Do floats allow enough precision for this? */ +      ( +         ((float) ZoO_index_random()) +         / ((float) ZoO_INDEX_MAX) +      ) +      * ((float) max) +   ); +} diff --git a/src/core/index.h b/src/core/index.h index 76e3507..1417662 100644 --- a/src/core/index.h +++ b/src/core/index.h @@ -3,6 +3,17 @@  #include "index_types.h" +/* + * Returns a random ZoO_index. + */ +ZoO_index ZoO_index_random (void); + +/* + * Returns a random ZoO_index, included in [0, limit] + */ +/*@ + @ ensures (\result <= limit); + @*/  ZoO_index ZoO_index_random_up_to (const ZoO_index limit);  #endif diff --git a/src/core/index_types.h b/src/core/index_types.h index 2d769ca..ad56d52 100644 --- a/src/core/index_types.h +++ b/src/core/index_types.h @@ -1,8 +1,10 @@  #ifndef _ZoO_CORE_INDEX_TYPES_H_  #define _ZoO_CORE_INDEX_TYPES_H_ +/* Must be unsigned. */  typedef unsigned int ZoO_index; +/* Must be > 0. */  #define ZoO_INDEX_MAX UINT_MAX  #endif diff --git a/src/core/knowledge.c b/src/core/knowledge.c deleted file mode 100644 index 94d76cd..0000000 --- a/src/core/knowledge.c +++ /dev/null @@ -1,15 +0,0 @@ -#include <stdlib.h> -#include <string.h> -#include <stdint.h> /* defines SIZE_MAX */ - -#include "../io/error.h" - -#include "knowledge.h" - -/** Basic functions of the ZoO_knowledge structure ****************************/ -void ZoO_knowledge_initialize (struct ZoO_knowledge k [const static 1]) -{ -   k->words = (struct ZoO_knowledge_word *) NULL; -   k->words_length = 0; -   k->words_sorted = (ZoO_index *) NULL; -} diff --git a/src/core/knowledge.h b/src/core/knowledge.h deleted file mode 100644 index 057e436..0000000 --- a/src/core/knowledge.h +++ /dev/null @@ -1,101 +0,0 @@ -#ifndef _ZoO_CORE_KNOWLEDGE_H_ -#define _ZoO_CORE_KNOWLEDGE_H_ - -#include "../core/char_types.h" -#include "../core/index_types.h" - -#include "knowledge_types.h" - -void ZoO_knowledge_initialize (struct ZoO_knowledge k [const restrict static 1]); - -void ZoO_knowledge_finalize (struct ZoO_knowledge k [const restrict static 1]); - - -/* - * When returning 0: - *    {word} was either added to {k} or its representation in {k} has its - *    occurrences count increased. - *    {*result} indicates where {word} is in {k->words}. - * - * When returning -1: - *    Something went wrong when adding the occurrence of {word} to {k}. - *    {k} remains semantically unchanged. - *    {*result} may or may not have been altered. - */ -int ZoO_knowledge_learn -( -   struct ZoO_knowledge k [const static 1], -   const ZoO_char word [const restrict static 1], -   ZoO_index result [const restrict static 1] -); - -int ZoO_knowledge_learn_sequence -( -   struct ZoO_knowledge k [const static 1], -   const ZoO_index sequence [const restrict], -   const ZoO_index sequence_length -); - -int ZoO_knowledge_get_following_sequences_ref -( -   const struct ZoO_knowledge k [const static 1], -   const ZoO_index initial_word, -   const ZoO_index * restrict following_sequences_ref [const restrict static 1], -   const ZoO_index * restrict following_sequences_weights [const restrict static 1], -   ZoO_index following_sequences_weights_sum [const static 1] -); - -int ZoO_knowledge_get_sequence -( -   const struct ZoO_knowledge k [const static 1], -   const ZoO_index sequences_ref, -   const ZoO_index * restrict sequence [const restrict static 1] -); - -int ZoO_knowledge_get_word -( -   const struct ZoO_knowledge k [const static 1], -   const ZoO_index word_ref, -   const ZoO_char * word [const restrict static 1], -   size_t word_size [const restrict static 1] -); - -/* - * When returning 0: - *    {word} is in {k}. - *    {word} is located at {k->words[*result]}. - * - * When returning -1: - *    {word} is not in {k}. - *    {*result} is where {word} was expected to be found in - *    {k->sorted_indices}. - */ -int ZoO_knowledge_find_word_id -( -   const struct ZoO_knowledge k [const restrict static 1], -   const ZoO_char word [const restrict static 1], -   ZoO_index result [const restrict static 1] -); - -int ZoO_knowledge_find_preceding_words -( -   const struct ZoO_knowledge k [const static 1], -   const ZoO_index sequence [const restrict], -   const ZoO_index markov_order, -   const ZoO_index * restrict preceding_words [const restrict static 1], -   const ZoO_index * restrict preceding_words_weights [const restrict static 1], -   ZoO_index preceding_words_weights_sum [const restrict static 1] -); - -int ZoO_knowledge_find_following_words -( -   const struct ZoO_knowledge k [const static 1], -   const ZoO_index sequence [const restrict], -   const ZoO_index sequence_length, -   const ZoO_index markov_order, -   const ZoO_index * restrict following_words [const restrict static 1], -   const ZoO_index * restrict following_words_weights [const restrict static 1], -   ZoO_index following_words_weights_sum [const restrict static 1] -); - -#endif diff --git a/src/core/knowledge_finalize.c b/src/core/knowledge_finalize.c deleted file mode 100644 index e4deda6..0000000 --- a/src/core/knowledge_finalize.c +++ /dev/null @@ -1,121 +0,0 @@ -#include <stdlib.h> -#include <string.h> -#include <stdint.h> /* defines SIZE_MAX */ - -#include "../io/error.h" - -#include "knowledge.h" - -void knowledge_sequence_collection_finalize -( -   struct ZoO_knowledge_sequence_collection c [const restrict static 1] -) -{ -   ZoO_index i; - -   if (c->sequences_ref != (ZoO_index *) NULL) -   { -      free((void *) c->sequences_ref); -      c->sequences_ref = (ZoO_index *) NULL; -   } - -   if (c->sequences_ref_sorted != (ZoO_index *) NULL) -   { -      free((void *) c->sequences_ref_sorted); -      c->sequences_ref_sorted = (ZoO_index *) NULL; -   } - -   if (c->occurrences != (ZoO_index *) NULL) -   { -      free((void *) c->occurrences); -      c->occurrences = (ZoO_index *) NULL; -   } - -   for (i = 0; i < c->sequences_ref_length; ++i) -   { -      free((void *) c->targets[i]); -      free((void *) c->targets_occurrences[i]); -   } - -   c->sequences_ref_length = 0; - -   if (c->targets != (ZoO_index **) NULL) -   { -      free((void *) c->targets); -      c->targets != (ZoO_index **) NULL; -   } - -   free((void *) c->targets_length); - -   if (c->targets_occurrences != (ZoO_index **) NULL) -   { -      free((void *) c->targets_occurrences); -      c->targets_occurrences != (ZoO_index **) NULL; -   } -} - -void knowledge_word_finalize -( -   struct ZoO_knowledge_word w [const restrict static 1] -) -{ -   w->word_size = 0; -   w->occurrences = 0; - -   if (w->word != (ZoO_char *) NULL) -   { -      free((void *) w->word); - -      w->word = (ZoO_char *) NULL; -   } - -   knowledge_sequence_collection_finalize(&(w->followed)); -   knowledge_sequence_collection_finalize(&(w->preceded)); -} - -void ZoO_knowledge_finalize (struct ZoO_knowledge k [const restrict static 1]) -{ -   ZoO_index i; - -   for (i = 0; i < k->words_length; ++i) -   { -      knowledge_word_finalize(k->words + i); -   } - -   k->words_length = 0; - -   if (k->words != (struct ZoO_knowledge_word *) NULL) -   { -      free((void *) k->words); - -      k->words = (struct ZoO_knowledge_word *) NULL; -   } - -   if (k->words_sorted != (ZoO_index *) NULL) -   { -      free((void *) k->words_sorted); - -      k->words_sorted = (ZoO_index *) NULL; -   } - -   for (i = 0; i < k->sequences_length; ++i) -   { -      free((void *) k->sequences[i]); -   } - -   k->sequences_length = 0; - -   if (k->sequences != (ZoO_index **) NULL) -   { -      free((void *) k->sequences); - -      k->sequences = (ZoO_index **) NULL; -   } - -   if (k->sequences_sorted != (ZoO_index *) NULL) -   { -      free((void *) k->sequences_sorted); - -      k->sequences_sorted = (ZoO_index *) NULL; -   } -} diff --git a/src/core/knowledge_search.c b/src/core/knowledge_search.c deleted file mode 100644 index d0c61ef..0000000 --- a/src/core/knowledge_search.c +++ /dev/null @@ -1,335 +0,0 @@ -#include <stdlib.h> - -#include "../core/char.h" -#include "../core/index.h" -#include "../core/sequence.h" - -#include "../io/error.h" - -#include "knowledge.h" - -/* See "knowledge.h". */ -int ZoO_knowledge_find_word_id -( -   const struct ZoO_knowledge k [const restrict static 1], -   const ZoO_char word [const restrict static 1], -   ZoO_index result [const restrict static 1] -) -{ -   /* This is a binary search */ -   int cmp; -   ZoO_index i, current_min, current_max; -   ZoO_index candidate_id; - -   /* Handles the case where the list is empty ********************************/ -   current_max = k->words_length; - -   if (current_max == 0) -   { -      *result = 0; - -      return -1; -   } -   /***************************************************************************/ - -   current_min = 0; -   current_max -= 1; - -   for (;;) -   { -      i = (current_min + ((current_max - current_min) / 2)); - -      cmp = ZoO_word_cmp(word, k->words[k->words_sorted[i]].word); - -      if (cmp > 0) -      { -         current_min = (i + 1); - -         if (current_min > current_max) -         { -            *result = current_min; - -            return -1; -         } -      } -      else if (cmp < 0) -      { -         if ((current_min > current_max) || (i == 0)) -         { -            *result = current_min; - -            return -1; -         } - -         current_max = (i - 1); -      } -      else -      { -         *result = i; - -         return 0; -      } -   } -} - -int ZoO_knowledge_find_preceding_words -( -   const struct ZoO_knowledge k [const static 1], -   const ZoO_index sequence [const restrict], -   const ZoO_index markov_order, /* Pre: (> 0) */ -   const ZoO_index * restrict preceding_words [const restrict static 1], -   const ZoO_index * restrict preceding_words_weights [const restrict static 1], -   ZoO_index preceding_words_weights_sum [const restrict static 1] -) -{ -   /* This is a binary search */ -   int cmp; -   ZoO_index i, current_min, current_max, local_sequence; -   const ZoO_index * restrict candidate; -   const ZoO_index markov_sequence_length = (markov_order - 1); -   const ZoO_index word = sequence[markov_sequence_length]; - -   if (word >= k->words_length) -   { -      ZoO_S_ERROR -      ( -         "Attempting to find the preceding words of an unknown word." -      ); - -      *preceding_words = (const ZoO_index *) NULL; -      *preceding_words_weights = (const ZoO_index *) NULL; -      *preceding_words_weights_sum = 0; - -      return -1; -   } - - -   if (markov_order == 1) -   { -      /* Special case: empty sequences. */ -      *preceding_words = (const ZoO_index *) k->words[word].preceded.targets; - -      *preceding_words_weights = -         (const ZoO_index *) k->words[word].preceded.targets_occurrences; - -      *preceding_words_weights_sum = k->words[word].occurrences; - -      return 0; -   } - -   /* Handles the case where the list is empty ********************************/ -   current_max = k->words[word].preceded.sequences_ref_length; - -   if (current_max == 0) -   { -      *preceding_words = (const ZoO_index *) NULL; -      *preceding_words_weights = (const ZoO_index *) NULL; -      *preceding_words_weights_sum = 0; - -      ZoO_S_ERROR -      ( -         "Attempting to find the preceding words of a sequence that never had " -         "any." -      ); - -      return -2; -   } -   /***************************************************************************/ - -   current_min = 0; -   current_max -= 1; - -   for (;;) -   { -      i = (current_min + ((current_max - current_min) / 2)); - -      local_sequence = k->words[word].preceded.sequences_ref_sorted[i]; - -      (void) ZoO_knowledge_get_sequence -      ( -         k, -         k->words[word].preceded.sequences_ref[local_sequence], -         &candidate -      ); - -      cmp = -         ZoO_sequence_cmp -         ( -            sequence, -            markov_sequence_length, -            candidate, -            markov_sequence_length -         ); - -      if (cmp > 0) -      { -         current_min = (i + 1); - -         if (current_min > current_max) -         { -            *preceding_words = (const ZoO_index *) NULL; -            *preceding_words_weights = (const ZoO_index *) NULL; -            *preceding_words_weights_sum = 0; - -            return -2; -         } -      } -      else if (cmp < 0) -      { -         if ((current_min > current_max) || (i == 0)) -         { -            *preceding_words = (const ZoO_index *) NULL; -            *preceding_words_weights = (const ZoO_index *) NULL; -            *preceding_words_weights_sum = 0; - -            return -2; -         } - -         current_max = (i - 1); -      } -      else -      { -         *preceding_words = k->words[word].preceded.targets[local_sequence]; - -         *preceding_words_weights = -            k->words[word].preceded.targets_occurrences[local_sequence]; - -         *preceding_words_weights_sum = -            k->words[word].preceded.occurrences[local_sequence]; - -         return 0; -      } -   } -} - -int ZoO_knowledge_find_following_words -( -   const struct ZoO_knowledge k [const static 1], -   const ZoO_index sequence [const restrict], -   const ZoO_index sequence_length, -   const ZoO_index markov_order, -   const ZoO_index * restrict following_words [const restrict static 1], -   const ZoO_index * restrict following_words_weights [const restrict static 1], -   ZoO_index following_words_weights_sum [const restrict static 1] -) -{ -   /* This is a binary search */ -   int cmp; -   ZoO_index i, current_min, current_max, local_sequence; -   const ZoO_index * restrict candidate; -   const ZoO_index markov_sequence_length = (markov_order - 1); -   const ZoO_index sequence_offset = -      ((sequence_length - markov_sequence_length) - 1); -   const ZoO_index word = sequence[sequence_offset]; - -   if (word >= k->words_length) -   { -      ZoO_S_ERROR -      ( -         "Attempting to find the following words of an unknown word." -      ); - -      *following_words = (const ZoO_index *) NULL; -      *following_words_weights = (const ZoO_index *) NULL; -      *following_words_weights_sum = 0; - -      return -1; -   } - -   if (markov_order == 1) -   { -      /* Special case: empty sequences. */ -      *following_words = (const ZoO_index *) k->words[word].preceded.targets; - -      *following_words_weights = -         (const ZoO_index *) k->words[word].preceded.targets_occurrences; - -      *following_words_weights_sum = k->words[word].occurrences; - -      return 0; -   } - -   /* Handles the case where the list is empty ********************************/ -   current_max = k->words[word].preceded.sequences_ref_length; - -   if (current_max == 0) -   { -      *following_words = (const ZoO_index *) NULL; -      *following_words_weights = (const ZoO_index *) NULL; -      *following_words_weights_sum = 0; - -      ZoO_S_WARNING -      ( -         "Attempting to find the following words of a sequence that never had " -         "any." -      ); - -      return -2; -   } -   /***************************************************************************/ - -   current_min = 0; -   current_max -= 1; - -   for (;;) -   { -      i = (current_min + ((current_max - current_min) / 2)); - -      local_sequence = k->words[word].followed.sequences_ref_sorted[i]; - -      (void) ZoO_knowledge_get_sequence -      ( -         k, -         k->words[word].followed.sequences_ref[local_sequence], -         &candidate -      ); - -      cmp = -         ZoO_sequence_cmp -         ( -            (sequence + sequence_offset), -            markov_sequence_length, -            candidate, -            markov_sequence_length -         ); - -      if (cmp > 0) -      { -         current_min = (i + 1); - -         if (current_min > current_max) -         { -            *following_words = (const ZoO_index *) NULL; -            *following_words_weights = (const ZoO_index *) NULL; -            *following_words_weights_sum = 0; - -            return -2; -         } -      } -      else if (cmp < 0) -      { -         if ((current_min > current_max) || (i == 0)) -         { -            *following_words = (const ZoO_index *) NULL; -            *following_words_weights = (const ZoO_index *) NULL; -            *following_words_weights_sum = 0; - -            return -2; -         } - -         current_max = (i - 1); -      } -      else -      { -         *following_words = k->words[word].followed.targets[local_sequence]; - -         *following_words_weights = -            k->words[word].followed.targets_occurrences[local_sequence]; - -         *following_words_weights_sum = -            k->words[word].followed.occurrences[local_sequence]; - -         return 0; -      } -   } -} diff --git a/src/core/knowledge_types.h b/src/core/knowledge_types.h deleted file mode 100644 index acd239f..0000000 --- a/src/core/knowledge_types.h +++ /dev/null @@ -1,37 +0,0 @@ -#ifndef _ZoO_CORE_KNOWLEDGE_TYPES_H_ -#define _ZoO_CORE_KNOWLEDGE_TYPES_H_ - -#include "../core/index_types.h" -#include "../core/char_types.h" - -struct ZoO_knowledge_sequence_collection -{ -   ZoO_index * sequences_ref; -   ZoO_index sequences_ref_length; -   ZoO_index * sequences_ref_sorted; -   ZoO_index * occurrences; -   ZoO_index ** targets; -   ZoO_index * targets_length; -   ZoO_index ** targets_occurrences; -}; - -struct ZoO_knowledge_word -{ -   const ZoO_char * word; -   size_t word_size; -   ZoO_index occurrences; -   struct ZoO_knowledge_sequence_collection followed; -   struct ZoO_knowledge_sequence_collection preceded; -}; - -struct ZoO_knowledge -{ -   struct ZoO_knowledge_word * words; -   ZoO_index words_length; -   ZoO_index * words_sorted; -   ZoO_index ** sequences; -   ZoO_index sequences_length; -   ZoO_index * sequences_sorted; -}; - -#endif diff --git a/src/core/main.c b/src/core/main.c deleted file mode 100644 index bb4ae23..0000000 --- a/src/core/main.c +++ /dev/null @@ -1,436 +0,0 @@ -#include <stdlib.h> -#include <stdio.h> -#include <string.h> -#include <time.h> -#include <signal.h> - -#include "../tool/strings.h" - -#include "../io/error.h" -#include "../io/parameters.h" -#include "../io/data_input.h" -#include "../io/data_output.h" -#include "../io/network.h" - -#include "knowledge.h" - -#include "state_types.h" - -static int run = 1; - -static void request_termination (int const signo) -{ -   if ((signo == SIGINT) || (signo == SIGTERM)) -   { -      run = 0; -   } -} - -static int initialize -( -   struct ZoO_state s [const static 1], -   int const argc, -   const char * argv [const static argc] -) -{ -   ZoO_S_DEBUG(ZoO_DEBUG_PROGRAM_FLOW, "Zero of One is initializing..."); - -   srand(time(NULL)); - -   /* prevents s [restrict] */ -   if (ZoO_knowledge_initialize(&(s->knowledge)) < 0) -   { -      return -1; -   } - -   if (ZoO_parameters_initialize(&(s->param), argc, argv) < 1) -   { -      ZoO_knowledge_finalize(&(s->knowledge)); - -      return -1; -   } - -   return 0; -} - -static int load_data_file (struct ZoO_state s [const static 1]) -{ -   struct ZoO_data_input input; -   char * result; - -   if (ZoO_data_input_open(&input, s->param.data_filename) < 0) -   { -      return -1; -   } - -   while -   ( -      ZoO_data_input_read_line -      ( -         &input, -         ZoO_knowledge_punctuation_chars_count, -         ZoO_knowledge_punctuation_chars -      ) == 0 -   ) -   { -      (void) ZoO_knowledge_assimilate -      ( -         &(s->knowledge), -         &(input.string), -         s->param.aliases_count, -         s->param.aliases -      ); -   } - -   ZoO_data_input_close(&input); - -   return 0; -} - -static int finalize (struct ZoO_state s [const static 1]) -{ -   int error; - -   ZoO_S_DEBUG(ZoO_DEBUG_PROGRAM_FLOW, "Zero of One is finalizing..."); - -   error = 0; - -   /* prevents s [restrict] */ -   ZoO_knowledge_finalize(&(s->knowledge)); - -   return error; -} - -static int network_connect (struct ZoO_state s [const static 1]) -{ -   return -      ZoO_network_connect -      ( -         &(s->network), -         s->param.irc_server_addr, -         s->param.irc_server_port, -         s->param.irc_server_channel, -         s->param.irc_username, -         s->param.irc_realname, -         s->param.aliases[0] -      ); -} - -static int should_reply -( -   struct ZoO_parameters param [const restrict static 1], -   struct ZoO_strings string [const restrict static 1], -   int should_learn [const restrict static 1] -) -{ -   ZoO_index i, j; - -   for (i = 0; i < param->aliases_count; ++i) -   { -      if (ZoO_IS_PREFIX(param->aliases[i], string->words[0])) -      { -         *should_learn = 0; - -         return 1; -      } - -      for (j = 1; j < string->words_count; ++j) -      { -         if (ZoO_IS_PREFIX(param->aliases[i], string->words[j])) -         { -            *should_learn = 1; - -            return 1; -         } -      } -   } - -   *should_learn = 1; - -   return (param->reply_rate >= (rand() % 100)); -} - -static void handle_user_join -( -   struct ZoO_state s [const static 1], -   struct ZoO_strings string [const restrict static 1], -   ssize_t const msg_offset, -   ssize_t const msg_size -) -{ -   ZoO_char * line; -   ZoO_index loc; - -   if (s->param.reply_rate < (rand() % 100)) -   { -      return; -   } - -   if -   ( -      ZoO_strings_parse -      ( -         string, -         (size_t) msg_size, -         (s->network.in + msg_offset), -         ZoO_knowledge_punctuation_chars_count, -         ZoO_knowledge_punctuation_chars -      ) < 0 -   ) -   { -      ZoO_S_DEBUG(ZoO_DEBUG_PROGRAM_FLOW, "Could not dissect join username."); - -      return; -   } - -   if -   ( -      ( -      ZoO_knowledge_find -         ( -            &(s->knowledge), -            string->words[0], -            &loc -         ) < 0 -      ) -      || (s->knowledge.words[loc].backward_links_count <= 3) -      || (s->knowledge.words[loc].forward_links_count <= 3) -   ) -   { -      if -      ( -         ZoO_knowledge_extend -         ( -            &(s->knowledge), -            (struct ZoO_strings *) NULL, -            0, -            (const char **) NULL, -            &line -         ) == 0 -      ) -      { -         if (line[0] == ' ') -         { -            strcpy((s->network.out), (line + 1)); -         } -         else -         { -            strcpy((s->network.out), line); -         } - -         free((void *) line); - -         ZoO_network_send(&(s->network)); -      } -   } -   else -   { -      if -      ( -         ZoO_knowledge_extend -         ( -            &(s->knowledge), -            string, -            0, -            (const char **) NULL, -            &line -         ) == 0 -      ) -      { -         if (line[0] == ' ') -         { -            strcpy((s->network.out), (line + 1)); -         } -         else -         { -            strcpy((s->network.out), line); -         } - -         free((void *) line); - -         ZoO_network_send(&(s->network)); -      } -   } -} - -static void handle_message -( -   struct ZoO_state s [const static 1], -   struct ZoO_strings string [const restrict static 1], -   ssize_t const msg_offset, -   /* FIXME: somehow we end up using (msg_size + 1), meaning there's a mixup -    *        between size and length. -    */ -   ssize_t const msg_size -) -{ -   ZoO_char * line; -   int reply, learn; - -   if -   ( -      ZoO_strings_parse -      ( -         string, -         (size_t) msg_size, -         (s->network.in + msg_offset), -         ZoO_knowledge_punctuation_chars_count, -         ZoO_knowledge_punctuation_chars -      ) < 0 -   ) -   { -      ZoO_S_DEBUG(ZoO_DEBUG_PROGRAM_FLOW, "Could not dissect msg."); - -      return; -   } - -   if (string->words_count == 0) -   { -      return; -   } - -   reply = should_reply(&(s->param), string, &learn); - -   if (learn) -   { -      /* -       * It would be best to do that after replying, but by then we no longer -       * have the string in 's->network.in'. -       */ -      (void) ZoO_data_output_write_line -      ( -         s->param.new_data_filename, -         (s->network.in + msg_offset), -         (size_t) (msg_size + 1) -      ); -   } - -   if -   ( -      reply -      && -      ( -         ZoO_knowledge_extend -         ( -            &(s->knowledge), -            string, -            s->param.aliases_count, -            s->param.aliases, -            &line -         ) == 0 -      ) -   ) -   { -      if (line[0] == ' ') -      { -         strcpy((s->network.out), (line + 1)); -      } -      else -      { -         strcpy((s->network.out), line); -      } - -      free((void *) line); - -      ZoO_network_send(&(s->network)); -   } - -   if (learn) -   { -      (void) ZoO_knowledge_assimilate -      ( -         &(s->knowledge), -         string, -         s->param.aliases_count, -         s->param.aliases -      ); -   } -} - -static int main_loop (struct ZoO_state s [const static 1]) -{ -   struct ZoO_strings string; -   ssize_t msg_offset, msg_size; -   enum ZoO_msg_type msg_type; - -   msg_offset = 0; -   msg_size = 0; - -   ZoO_strings_initialize(&string); - -   while (run) -   { -      if -      ( -         ZoO_network_receive -         ( -            &(s->network), -            &msg_offset, -            &msg_size, -            &msg_type -         ) == 0 -      ) -      { -         switch (msg_type) -         { -            case ZoO_JOIN: -               handle_user_join(s, &string, msg_offset, msg_size); -               break; - -            case ZoO_PRIVMSG: -               handle_message(s, &string, msg_offset, msg_size); -               break; -         } -      } -   } - -   ZoO_strings_finalize(&string); - -   ZoO_network_disconnect(&(s->network)); - -   return 0; -} - -int main (int const argc, const char * argv [const static argc]) -{ -   struct ZoO_state s; - -   if (initialize(&s, argc, argv) < 0) -   { -      return -1; -   } - -   if (load_data_file(&s) < 0) -   { -      goto CRASH; -   } - -   if (network_connect(&s) < 0) -   { -      goto CRASH; -   } - -   if (main_loop(&s) < 0) -   { -      goto CRASH; -   } - -   (void) finalize(&s); - -   ZoO_S_DEBUG(ZoO_DEBUG_PROGRAM_FLOW, "Zero of One terminated normally."); - -   return 0; - -   CRASH: -   { -      (void) finalize(&s); - -      ZoO_S_DEBUG -      ( -         ZoO_DEBUG_PROGRAM_FLOW, -         "Zero of One terminated by crashing." -      ); - -      return -1; -   } -} diff --git a/src/core/sequence.c b/src/core/sequence.c index 9e370a3..d7ff9d0 100644 --- a/src/core/sequence.c +++ b/src/core/sequence.c @@ -5,18 +5,56 @@  #include "sequence.h" +/* + * Bypass rendundant ZoO_START_OF_SEQUENCE_ID at the start of a sequence. + */ +/* ensures (*sequence_offset <= sequence_length) */ +static void bypass_redundant_sos +( +   const ZoO_index sequence [const restrict], +   const ZoO_index sequence_length, +   ZoO_index sequence_offset [const restrict static 1] +) +{ +   ZoO_index i; + +   *sequence_offset = 0; + +   for (i = 0; i < sequence_length; ++i) +   { +      if (sequence[i] != ZoO_START_OF_SEQUENCE_ID) +      { +         return; +      } +      else if (sequence[i] == ZoO_START_OF_SEQUENCE_ID) +      { +         *sequence_offset = i; +      } +   } +} + +  /* See "sequence.h" */  int ZoO_sequence_cmp  (     const ZoO_index sequence_a [const], -   const ZoO_index sequence_a_length, +   ZoO_index sequence_a_length,     const ZoO_index sequence_b [const], -   const ZoO_index sequence_b_length +   ZoO_index sequence_b_length  )  { -   ZoO_index min_length; +   ZoO_index min_length, a, b; +   ZoO_index a_offset, b_offset;     ZoO_index i; +   bypass_redundant_sos(sequence_a, sequence_a_length, &a_offset); +   bypass_redundant_sos(sequence_b, sequence_b_length, &b_offset); + +   /*@ requires (*a_offset <= sequence_a_length) @*/ +   sequence_a_length -= a_offset; +   /*@ requires (*b_offset <= sequence_b_length) @*/ +   sequence_b_length -= b_offset; +     if (sequence_a_length < sequence_b_length)     {        min_length = sequence_a_length; @@ -26,47 +64,37 @@ int ZoO_sequence_cmp        min_length = sequence_b_length;     } +   /*@ ensures (min_length <= sequence_a_length) @*/ +   /*@ ensures (min_length <= sequence_b_length) @*/ +     for (i = 0; i < min_length; ++i)     { -      if (sequence_a[i] < sequence_b[i]) +      /*@ requires ((i + a_offset) < sequence_a_length) @*/ +      a = sequence_a[i + a_offset]; +      /*@ requires ((i + b_offset) < sequence_b_length) @*/ +      b = sequence_b[i + b_offset]; + +      if (a < b)        {           return -1;        } -      else if (sequence_b[i] > sequence_b[i]) +      else if (b > a)        {           return 1;        } -      else if -      ( -         (sequence_a[i] == ZoO_END_OF_SEQUENCE_ID) -         && (sequence_b[i] == ZoO_END_OF_SEQUENCE_ID) -      ) +      else if ((a == ZoO_END_OF_SEQUENCE_ID) && (b == ZoO_END_OF_SEQUENCE_ID))        {           return 0;        }     } -   if (sequence_a_length < sequence_b_length) +   if (sequence_a_length > sequence_b_length)     { -      if (sequence_b[i] == ZoO_END_OF_SEQUENCE_ID) -      { -         return 0; -      } -      else -      { -         return -1; -      } +      return 1;     } -   else if (sequence_a_length > sequence_b_length) +   else if (sequence_a_length < sequence_b_length)     { -      if (sequence_a[i] == ZoO_END_OF_SEQUENCE_ID) -      { -         return 0; -      } -      else -      { -         return 1; -      } +      return -1;     }     else     { diff --git a/src/core/sequence.h b/src/core/sequence.h index e609b4d..77ecd6c 100644 --- a/src/core/sequence.h +++ b/src/core/sequence.h @@ -1,11 +1,21 @@  #ifndef _ZoO_CORE_SEQUENCE_H_  #define _ZoO_CORE_SEQUENCE_H_ +#include "../core/char_types.h"  #include "../core/index_types.h" -#include "../core/knowledge_types.h" +#include "../knowledge/knowledge_types.h"  #include "sequence_types.h" +int ZoO_sequence_from_undercase_string +( +   const ZoO_char string [const restrict], +   const ZoO_index string_length, +   struct ZoO_knowledge k [const restrict static 1], +   ZoO_index * sequence [const restrict static 1], +   ZoO_index sequence_length [const restrict static 1] +); +  /*   * Creates a sequence containing {initial_word}. The remaining elements of   * sequence are added according to what is known to {k} as being possible. @@ -42,7 +52,13 @@ int ZoO_sequence_create_from   * ZoO_END_OF_SEQUENCE marks the ending of a sequence, regardless of indicated   * sequence length, meaning that [10][ZoO_END_OF_SEQUENCE][9] and   * [10][ZoO_END_OF_SEQUENCE][8] are considered equal. Sequences do not have to - * contain ZoO_END_OF_SEQUENCE. + * contain ZoO_END_OF_SEQUENCE. [10][ZoO_END_OF_SEQUENCE] and [10] are + * considered different, [10][ZoO_END_OF_SEQUENCE] + * and [10][ZoO_END_OF_SEQUENCE][ZoO_END_OF_SEQUENCE] are considered equal. + * Same logic is applyied for ZoO_START_OF_SEQUENCE: + * [START_OF_SEQUENCE][10] is not [10], but + * [START_OF_SEQUENCE][START_OF_SEQUENCE][10] and [START_OF_SEQUENCE][10] are + * the same.   * Return:   *    1 iff {sequence_a} should be considered being more than {sequence_b}   *    0 iff {sequence_a} should be considered being equal to {sequence_b} diff --git a/src/core/sequence_creation.c b/src/core/sequence_creation.c index 1133be9..f460629 100644 --- a/src/core/sequence_creation.c +++ b/src/core/sequence_creation.c @@ -19,6 +19,11 @@   *    (> weights_sum 0).   *    (= (sum weights) weights_sum).   */ +/*@ + @ requires (weights_sum > 0); + @ requires \valid(weights); + @ requires (\sum(0, (\length(weights) - 1), weights) = weights_sum); +@*/  static ZoO_index weighted_random_pick  (     const ZoO_index weights [const restrict static 1], @@ -29,12 +34,12 @@ static ZoO_index weighted_random_pick     accumulator = 0; -   /* Safe: Included in [0, weights_sum]. */     random_number = ZoO_index_random_up_to(weights_sum); +   /*@ ensures (0 <= random_number <= weights_sum); @*/     for (result = 0; accumulator < random_number; ++result)     { -      /* Safe: (= (sum weights) weights_sum) */ +      /*@ requires (\sum(0, (\length(weights) - 1), weights) = weights_sum); @*/        accumulator += weights[result];     } diff --git a/src/core/sequence_from_string.c b/src/core/sequence_from_string.c new file mode 100644 index 0000000..51d7049 --- /dev/null +++ b/src/core/sequence_from_string.c @@ -0,0 +1,315 @@ +#define _POSIX_C_SOURCE 200809L +#include <stdlib.h> +#include <string.h> +#include <stdint.h> /* defines SIZE_MAX */ + +#include "../core/char.h" +#include "../core/index.h" + +#include "../cli/cli.h" + +#include "../knowledge/knowledge.h" + +#include "sequence.h" + +static int add_word_id_to_sequence +( +   const ZoO_index word_id, +   ZoO_index * sequence [const restrict static 1], +   ZoO_index sequence_length [const restrict static 1] +) +{ +   ZoO_index * new_sequence; + +   *sequence_length += 1; + +   new_sequence = +      (ZoO_index *) realloc +      ( +         (void *) *sequence, +         (((size_t) sequence_length) * sizeof(ZoO_index)) +      ); + +   if (new_sequence == (ZoO_index *) NULL) +   { +      ZoO_S_ERROR("Unable to reallocate a sequence to add word ids to it."); + +      return -1; +   } + +   return 0; +} + +/******************************************************************************/ +/** HANDLING PUNCTUATION ******************************************************/ +/******************************************************************************/ +static int add_punctuation_to_sequence +( +   const ZoO_char string [const restrict static 1], +   const ZoO_char punctuation, +   ZoO_index * sequence [const restrict static 1], +   ZoO_index sequence_length [const restrict static 1], +   const struct ZoO_knowledge k [const restrict static 1] +) +{ +   ZoO_index word_id; +   ZoO_char as_word[2]; + +   as_word[0] = punctuation; +   as_word[1] = '\0'; + +   if (ZoO_knowledge_find_word_id(k, as_word, 2, &word_id) < 0) +   { +      ZoO_PROG_ERROR +      ( +         "'%s' was defined as a punctuation, was found in a string, yet is not" +         " defined in the knowledge database.", +         as_word +      ); + +      return -1; +   } + +   if (add_word_id_to_sequence(word_id, sequence, sequence_length) < 0) +   { +      return -1; +   } + +   return 0; +} + +static int word_is_punctuation_terminated +( +   const ZoO_char string [const restrict static 1], +   const ZoO_index word_start, +   const ZoO_index word_length +) +{ +   return ZoO_char_is_punctuation(string[word_length]); +} + +/******************************************************************************/ +/** HANDLING WORDS ************************************************************/ +/******************************************************************************/ +static int add_word_to_sequence +( +   const ZoO_char string [const restrict static 1], +   const ZoO_index word_start, +   const ZoO_index word_length, +   ZoO_index * sequence [const restrict static 1], +   ZoO_index sequence_length [const restrict static 1], +   struct ZoO_knowledge k [const restrict static 1] +) +{ +   ZoO_index word_id; +   ZoO_char * stored_word; + +   if (word_length == 0) +   { +      return 0; +   } + +   if +   ( +      ZoO_knowledge_learn_word +      ( +         k, +         (string + word_start), +         word_length, +         &word_id +      ) < 0 +   ) +   { +      return -1; +   } + +   if (add_word_id_to_sequence(word_id, sequence, sequence_length) < 0) +   { +      return -1; +   } + +   return 0; +} + +static int add_finding_to_sequence +( +   const ZoO_char string [const restrict static 1], +   const ZoO_index word_start, +   const ZoO_index word_length, +   ZoO_index * sequence [const restrict static 1], +   ZoO_index sequence_length [const restrict static 1], +   struct ZoO_knowledge k [const restrict static 1] +) +{ +   ZoO_index punctuation; + +   if (word_is_punctuation_terminated(string, word_start, word_length)) +   { +      punctuation = 1; +   } +   else +   { +      punctuation = 0; +   } + +   if +   ( +      add_word_to_sequence +      ( +         string, +         word_start, +         (word_length - punctuation), +         sequence, +         sequence_length, +         k +      ) < 0 +   ) +   { +      return -1; +   } + +   if +   ( +      (punctuation == 1) +      && +      ( +         add_punctuation_to_sequence +         ( +            string, +            string[word_start + word_length - 1], +            sequence, +            sequence_length, +            k +         ) < 0 +      ) +   ) +   { +      return -1; +   } + +   return 0; +} + +static int find_word +( +   const ZoO_char string [const restrict static 1], +   const ZoO_index string_length, +   const ZoO_index offset, +   ZoO_index word_start [const restrict static 1], +   ZoO_index word_length [const restrict static 1] +) +{ +   ZoO_index i; + +   i = offset; + +   while ((string[i] == ' ') && (i < string_length)) +   { +      i += 1; +   } + +   if (i >= string_length) +   { +      return -1; +   } + +   *word_start = i; + +   while ((string[i] != ' ') && (i < string_length)) +   { +      i += 1; +   } + +   if (i >= string_length) +   { +      return -1; +   } + +   *word_length = (i - *word_start); + +   return 0; +} + +/******************************************************************************/ +/** EXPORTED ******************************************************************/ +/******************************************************************************/ +int ZoO_sequence_from_undercase_string +( +   const ZoO_char string [const restrict], +   const ZoO_index string_length, +   struct ZoO_knowledge k [const restrict static 1], +   ZoO_index * sequence [const restrict static 1], +   ZoO_index sequence_length [const restrict static 1] +) +{ +   ZoO_index word_start, word_length; +   ZoO_index i; + +   i = 0; + +   *sequence = (ZoO_index *) NULL; +   *sequence_length = 0; + +   if +   ( +      add_word_id_to_sequence +      ( +         ZoO_START_OF_SEQUENCE_ID, +         sequence, +         sequence_length +      ) < 0 +   ) +   { +      return -1; +   } + +   while (i < string_length) +   { +      if (find_word(string, i, string_length, &word_start, &word_length) < 0) +      { +         break; +      } + +      if +      ( +         add_finding_to_sequence +         ( +            string, +            word_start, +            word_length, +            sequence, +            sequence_length, +            k +         ) < 0 +      ) +      { +         free((void *) *sequence); +         *sequence = (ZoO_index *) NULL; +         *sequence_length = 0; + +         return -1; +      } + +      i = (word_start + word_length); +   } + +   if +   ( +      add_word_id_to_sequence +      ( +         ZoO_END_OF_SEQUENCE_ID, +         sequence, +         sequence_length +      ) < 0 +   ) +   { +      free((void *) *sequence); + +      *sequence = (ZoO_index *) NULL; +      *sequence_length = 0; + +      return -1; +   } + +   return 0; +} diff --git a/src/core/sequence_types.h b/src/core/sequence_types.h index 717d418..c260a8a 100644 --- a/src/core/sequence_types.h +++ b/src/core/sequence_types.h @@ -3,7 +3,8 @@  #define ZoO_START_OF_SEQUENCE_ID 0  #define ZoO_END_OF_SEQUENCE_ID   1 +#define ZoO_ACTION_SEQUENCE_ID   2 -#define ZoO_RESERVED_IDS_COUNT   2 +#define ZoO_RESERVED_IDS_COUNT   3  #endif | 


