hermet pushed a commit to branch master. http://git.enlightenment.org/tools/enventor.git/commit/?id=fb7224f4a928c29935dfbd76825e1638c39d5c86
commit fb7224f4a928c29935dfbd76825e1638c39d5c86 Author: Hermet Park <[email protected]> Date: Mon Jul 4 12:58:07 2016 +0900 auto_comp: code refactoring. added descriptive comments for the main logic of context analysis. --- src/lib/auto_comp.c | 126 +++++++++++++++++++++++++++++++--------------------- 1 file changed, 76 insertions(+), 50 deletions(-) diff --git a/src/lib/auto_comp.c b/src/lib/auto_comp.c index 44e3560..0ff3131 100644 --- a/src/lib/auto_comp.c +++ b/src/lib/auto_comp.c @@ -8,7 +8,7 @@ #define QUEUE_SIZE 20 #define COMPSET_PAIR_MINIMUM 1 #define MAX_CONTEXT_STACK 40 -#define MAX_KEYWORD_LENGHT 40 +#define MAX_KEYWORD_LENGTH 40 typedef struct lexem_s { @@ -127,49 +127,46 @@ lexem_tree_free(lexem **root) static void context_lexem_thread_cb(void *data, Ecore_Thread *thread EINA_UNUSED) { + const int quot_len = QUOT_UTF8_LEN; + ctx_lexem_td *td = (ctx_lexem_td *)data; + if (!td->utf8) return; - Eina_Bool find_flag = EINA_FALSE; - Eina_Bool dot_lex = EINA_FALSE; - Eina_List *l = NULL; - Eina_List *nodes = td->ad->lexem_root->nodes; - td->result = (lexem *)td->ad->lexem_root; int cur_pos = td->cur_pos; if (cur_pos <= 1) return; + Eina_List *l = NULL; + Eina_List *nodes = td->ad->lexem_root->nodes; + td->result = (lexem *)td->ad->lexem_root; int i = 0; - - if (!td->utf8) return; - + int k = 0; + int depth = 0; + int context_len = 0; + char stack[MAX_CONTEXT_STACK][MAX_KEYWORD_LENGTH]; char *utf8 = td->utf8; char *cur = utf8; char *end = cur + cur_pos; - char stack[MAX_CONTEXT_STACK][MAX_KEYWORD_LENGHT]; - int depth = 0; char *help_ptr = NULL; char *help_end_ptr = NULL; - const char *quot = QUOT_UTF8; - const int quot_len = QUOT_UTF8_LEN; - int quot_cnt = 0; - int context_len = 0; + Eina_Bool inside_quot = EINA_FALSE; + Eina_Bool find_flag = EINA_FALSE; + Eina_Bool dot_lex = EINA_FALSE; while (cur && cur <= end) { + //Check inside quote if ((cur != end) && (!strncmp(cur, quot, quot_len))) { - /*TODO: add exception for case '\"'*/ - quot_cnt++; - cur++; - continue; + /* TODO: add exception for case '\"' */ + inside_quot = !inside_quot; } - if (quot_cnt % 2) + if (inside_quot) { cur++; continue; } - //Check inside comment if (*cur == '/') { @@ -185,75 +182,101 @@ context_lexem_thread_cb(void *data, Ecore_Thread *thread EINA_UNUSED) } } + //Case 1. Find a context and store it. if (*cur == '{') { - for (help_end_ptr = cur; - !isalnum(*help_end_ptr); - help_end_ptr--); - for (help_ptr = help_end_ptr; - ((help_ptr >= utf8) && - ((isalnum(*help_ptr )) || (*help_ptr == '_'))); - help_ptr--); + //Skip non-alpha numberics. + help_end_ptr = cur; + while (!isalnum(*help_end_ptr)) + help_end_ptr--; + + help_ptr = help_end_ptr; + + //Figure out current context keyword + while ((help_ptr >= utf8) && + (isalnum(*help_ptr) || (*help_ptr == '_'))) + { + help_ptr--; + } + if (help_ptr != utf8) help_ptr++; + //Exceptional case for size. context_len = help_end_ptr - help_ptr + 1; - if (context_len >= MAX_KEYWORD_LENGHT) + if (context_len >= MAX_KEYWORD_LENGTH) { cur++; continue; } - memset(stack[depth], 0x0, MAX_KEYWORD_LENGHT); + //Store this context. strncpy(stack[depth], help_ptr, context_len); - depth++; - if (depth == MAX_CONTEXT_STACK) break; + if ((++depth) == MAX_CONTEXT_STACK) break; } + + //Case 2. Find a context and store it. if (*cur == '.') { Eina_Bool alpha_present = EINA_FALSE; help_end_ptr = cur - 1; - for (help_ptr = help_end_ptr; help_ptr && isalnum(*help_ptr); help_ptr--) - if (isalpha(*help_ptr)) alpha_present = EINA_TRUE; + + //Backtrace to the beginning of a certain keyword. + for (help_ptr = help_end_ptr; help_ptr && isalnum(*help_ptr); + help_ptr--) + { + if (isalpha(*help_ptr)) alpha_present = EINA_TRUE; + } + + //Hanlding Exception cases. if ((!alpha_present) || (!strncmp(help_ptr, quot, quot_len))) { cur++; continue; } + if (help_ptr != utf8) help_ptr++; + context_len = help_end_ptr - help_ptr + 1; - if (context_len >= MAX_KEYWORD_LENGHT) + if (context_len >= MAX_KEYWORD_LENGTH) { cur++; continue; } - memset(stack[depth], 0x0, MAX_KEYWORD_LENGHT); + + //Store this context. + strncpy(stack[depth], help_ptr, context_len); strncpy(stack[depth], help_ptr, context_len); - depth++; - if (depth == MAX_CONTEXT_STACK) break; + if ((++depth) == MAX_CONTEXT_STACK) break; + dot_lex = EINA_TRUE; - } - if ((*cur == ';') && dot_lex) + } + + //End of a dot lex context. + //Reset the previous context if its out of scope. + if (dot_lex && (*cur == ';')) { dot_lex = EINA_FALSE; - memset(stack[depth], 0x0, MAX_KEYWORD_LENGHT); + memset(stack[depth], 0x0, MAX_KEYWORD_LENGTH); if (depth > 0) depth--; } + + //End of a context. Reset the previous context if its out of scope. if (*cur == '}') { - memset(stack[depth], 0x0, MAX_KEYWORD_LENGHT); + memset(stack[depth], 0x0, MAX_KEYWORD_LENGTH); if (depth > 0) depth--; } cur++; } - if (quot_cnt % 2) + if (inside_quot) { td->result = NULL; return; } - int k = 0; + // Find current context where a cursor is inside of among the stack. for (i = 0; i < depth; i++) { find_flag = EINA_FALSE; @@ -261,12 +284,15 @@ context_lexem_thread_cb(void *data, Ecore_Thread *thread EINA_UNUSED) { for (k = 0; k < td->result->name_count; k++) { - if (!strncmp(stack[i], td->result->name[k], strlen(td->result->name[k]))) - { - nodes = td->result->nodes; - find_flag = EINA_TRUE; - } - if (find_flag) break; + //Ok, We go through this context. + //FIXME: Need to compare elaborately? + if (!strncmp(stack[i], td->result->name[k], + strlen(td->result->name[k]))) + { + nodes = td->result->nodes; + find_flag = EINA_TRUE; + break; + } } if (find_flag) break; } --
