naoa
null+****@clear*****
Sun Aug 10 06:56:57 JST 2014
naoa 2014-08-10 06:56:57 +0900 (Sun, 10 Aug 2014) New Revision: be849e2ec03c14f0a7ae74ccdf388738c2f58524 https://github.com/groonga/groonga/commit/be849e2ec03c14f0a7ae74ccdf388738c2f58524 Merged c514768: Merge pull request #185 from naoa/add-highlight_html-function Message: Add highlight_html() function It tags output text after it html_escape and normalize text. It can use easily without option. highlight_html(column) TODO: Document it. Added files: test/command/suite/select/function/highlight_html/empty_column.expected test/command/suite/select/function/highlight_html/empty_column.test test/command/suite/select/function/highlight_html/empty_query.expected test/command/suite/select/function/highlight_html/empty_query.test test/command/suite/select/function/highlight_html/no_hit.expected test/command/suite/select/function/highlight_html/no_hit.test test/command/suite/select/function/highlight_html/no_query.expected test/command/suite/select/function/highlight_html/no_query.test test/command/suite/select/function/highlight_html/one_hit.expected test/command/suite/select/function/highlight_html/one_hit.test test/command/suite/select/function/highlight_html/one_keyword.expected test/command/suite/select/function/highlight_html/one_keyword.test test/command/suite/select/function/highlight_html/twice_hits.expected test/command/suite/select/function/highlight_html/twice_hits.test test/command/suite/select/function/highlight_html/two_keywords.expected test/command/suite/select/function/highlight_html/two_keywords.test Modified files: lib/proc.c Modified: lib/proc.c (+86 -0) =================================================================== --- lib/proc.c 2014-08-10 06:45:26 +0900 (5eaacaf) +++ lib/proc.c 2014-08-10 06:56:57 +0900 (bc13a42) @@ -4785,6 +4785,89 @@ grn_pat_tag_keys(grn_ctx *ctx, grn_obj *keywords, } static grn_obj * +func_highlight_html(grn_ctx *ctx, int nargs, grn_obj **args, + grn_user_data *user_data) +{ + grn_obj *highlighted = NULL; + +#define N_REQUIRED_ARGS 1 + if (nargs == N_REQUIRED_ARGS) { + grn_obj *string = args[0]; + grn_obj *expression = NULL; + grn_obj *condition_ptr = NULL; + grn_obj *condition = NULL; + grn_bool html_escape_flag = GRN_TRUE; + unsigned int n_keyword_sets = 1; + const char *open_tags[1]; + unsigned int open_tag_lengths[1]; + const char *close_tags[1]; + unsigned int close_tag_lengths[1]; + grn_obj *keywords; + grn_obj got_keywords; + + open_tags[0] = "<span class=\"keyword\">"; + open_tag_lengths[0] = strlen("<span class=\"keyword\">"); + close_tags[0] = "</span>"; + close_tag_lengths[0] = strlen("</span>"); + + keywords = grn_table_create(ctx, NULL, 0, NULL, + GRN_OBJ_TABLE_PAT_KEY, + grn_ctx_at(ctx, GRN_DB_SHORT_TEXT), + NULL); + { + grn_obj * normalizer; + normalizer = grn_ctx_get(ctx, "NormalizerAuto", -1); + grn_obj_set_info(ctx, keywords, GRN_INFO_NORMALIZER, normalizer); + grn_obj_unlink(ctx, normalizer); + } + + grn_proc_get_info(ctx, user_data, NULL, NULL, &expression); + condition_ptr = grn_expr_get_var(ctx, expression, + GRN_SELECT_INTERNAL_VAR_CONDITION, + strlen(GRN_SELECT_INTERNAL_VAR_CONDITION)); + if (condition_ptr) { + condition = GRN_PTR_VALUE(condition_ptr); + } + + if (condition) { + GRN_PTR_INIT(&got_keywords, GRN_OBJ_VECTOR, GRN_ID_NIL); + grn_expr_get_keywords(ctx, condition, &got_keywords); + + for (;;) { + grn_obj *keyword; + GRN_PTR_POP(&got_keywords, keyword); + if (!keyword) { break; } + grn_table_add(ctx, keywords, + GRN_TEXT_VALUE(keyword), + GRN_TEXT_LEN(keyword), + NULL); + } + grn_obj_unlink(ctx, &got_keywords); + } + + highlighted = GRN_PROC_ALLOC(GRN_DB_TEXT, 0); + grn_pat_tag_keys(ctx, keywords, + GRN_TEXT_VALUE(string), GRN_TEXT_LEN(string), + open_tags, + open_tag_lengths, + close_tags, + close_tag_lengths, + n_keyword_sets, + highlighted, + html_escape_flag); + + grn_obj_unlink(ctx, keywords); + } +#undef N_REQUIRED_ARGS + + if (!highlighted) { + highlighted = GRN_PROC_ALLOC(GRN_DB_VOID, 0); + } + + return highlighted; +} + +static grn_obj * func_highlight_full(grn_ctx *ctx, int nargs, grn_obj **args, grn_user_data *user_data) { @@ -5122,6 +5205,9 @@ grn_db_init_builtin_query(grn_ctx *ctx) grn_proc_set_selector(ctx, selector_proc, selector_between); } + grn_proc_create(ctx, "highlight_html", -1, GRN_PROC_FUNCTION, + func_highlight_html, NULL, NULL, 0, NULL); + grn_proc_create(ctx, "highlight_full", -1, GRN_PROC_FUNCTION, func_highlight_full, NULL, NULL, 0, NULL); } Added: test/command/suite/select/function/highlight_html/empty_column.expected (+15 -0) 100644 =================================================================== --- /dev/null +++ test/command/suite/select/function/highlight_html/empty_column.expected 2014-08-10 06:56:57 +0900 (3e75756) @@ -0,0 +1,15 @@ +table_create Entries TABLE_NO_KEY +[[0,0.0,0.0],true] +column_create Entries body COLUMN_SCALAR ShortText +[[0,0.0,0.0],true] +table_create Terms TABLE_PAT_KEY ShortText --default_tokenizer TokenBigram --normalizer NormalizerAuto +[[0,0.0,0.0],true] +column_create Terms document_index COLUMN_INDEX|WITH_POSITION Entries body +[[0,0.0,0.0],true] +load --table Entries +[ +{"body": ""} +] +[[0,0.0,0.0],1] +select Entries --output_columns --match_columns body --query 'groonga' --output_columns 'highlight_html(body)' --command_version 2 +[[0,0.0,0.0],[[[0],[["highlight_html","null"]]]]] Added: test/command/suite/select/function/highlight_html/empty_column.test (+14 -0) 100644 =================================================================== --- /dev/null +++ test/command/suite/select/function/highlight_html/empty_column.test 2014-08-10 06:56:57 +0900 (da7d660) @@ -0,0 +1,14 @@ +table_create Entries TABLE_NO_KEY +column_create Entries body COLUMN_SCALAR ShortText + +table_create Terms TABLE_PAT_KEY ShortText --default_tokenizer TokenBigram --normalizer NormalizerAuto +column_create Terms document_index COLUMN_INDEX|WITH_POSITION Entries body + +load --table Entries +[ +{"body": ""} +] + +select Entries --output_columns \ + --match_columns body --query 'groonga' \ + --output_columns 'highlight_html(body)' --command_version 2 Added: test/command/suite/select/function/highlight_html/empty_query.expected (+37 -0) 100644 =================================================================== --- /dev/null +++ test/command/suite/select/function/highlight_html/empty_query.expected 2014-08-10 06:56:57 +0900 (6734f65) @@ -0,0 +1,37 @@ +table_create Entries TABLE_NO_KEY +[[0,0.0,0.0],true] +column_create Entries body COLUMN_SCALAR ShortText +[[0,0.0,0.0],true] +table_create Terms TABLE_PAT_KEY ShortText --default_tokenizer TokenBigram --normalizer NormalizerAuto +[[0,0.0,0.0],true] +column_create Terms document_index COLUMN_INDEX|WITH_POSITION Entries body +[[0,0.0,0.0],true] +load --table Entries +[ +{"body": "Mroonga is a ＭｙＳＱＬ storage engine based on Groonga. <b>Rroonga</b> is a Ruby binding of Groonga."} +] +[[0,0.0,0.0],1] +select Entries --output_columns --match_columns body --query '' --output_columns 'highlight_html(body)' --command_version 2 +[ + [ + 0, + 0.0, + 0.0 + ], + [ + [ + [ + 1 + ], + [ + [ + "highlight_html", + "null" + ] + ], + [ + "Mroonga is a ＭｙＳＱＬ storage engine based on Groonga. <b>Rroonga</b> is a Ruby binding of Groonga." + ] + ] + ] +] Added: test/command/suite/select/function/highlight_html/empty_query.test (+14 -0) 100644 =================================================================== --- /dev/null +++ test/command/suite/select/function/highlight_html/empty_query.test 2014-08-10 06:56:57 +0900 (ab5e888) @@ -0,0 +1,14 @@ +table_create Entries TABLE_NO_KEY +column_create Entries body COLUMN_SCALAR ShortText + +table_create Terms TABLE_PAT_KEY ShortText --default_tokenizer TokenBigram --normalizer NormalizerAuto +column_create Terms document_index COLUMN_INDEX|WITH_POSITION Entries body + +load --table Entries +[ +{"body": "Mroonga is a ＭｙＳＱＬ storage engine based on Groonga. <b>Rroonga</b> is a Ruby binding of Groonga."} +] + +select Entries --output_columns \ + --match_columns body --query '' \ + --output_columns 'highlight_html(body)' --command_version 2 Added: test/command/suite/select/function/highlight_html/no_hit.expected (+15 -0) 100644 =================================================================== --- /dev/null +++ test/command/suite/select/function/highlight_html/no_hit.expected 2014-08-10 06:56:57 +0900 (eec979f) @@ -0,0 +1,15 @@ +table_create Entries TABLE_NO_KEY +[[0,0.0,0.0],true] +column_create Entries body COLUMN_SCALAR ShortText +[[0,0.0,0.0],true] +table_create Terms TABLE_PAT_KEY ShortText --default_tokenizer TokenBigram --normalizer NormalizerAuto +[[0,0.0,0.0],true] +column_create Terms document_index COLUMN_INDEX|WITH_POSITION Entries body +[[0,0.0,0.0],true] +load --table Entries +[ +{"body": "Mroonga is a ＭｙＳＱＬ storage engine based on Groonga. <b>Rroonga</b> is a Ruby binding of Groonga."} +] +[[0,0.0,0.0],1] +select Entries --output_columns --match_columns body --query 'droonga' --output_columns 'highlight_html(body)' --command_version 2 +[[0,0.0,0.0],[[[0],[["highlight_html","null"]]]]] Added: test/command/suite/select/function/highlight_html/no_hit.test (+14 -0) 100644 =================================================================== --- /dev/null +++ test/command/suite/select/function/highlight_html/no_hit.test 2014-08-10 06:56:57 +0900 (41700fb) @@ -0,0 +1,14 @@ +table_create Entries TABLE_NO_KEY +column_create Entries body COLUMN_SCALAR ShortText + +table_create Terms TABLE_PAT_KEY ShortText --default_tokenizer TokenBigram --normalizer NormalizerAuto +column_create Terms document_index COLUMN_INDEX|WITH_POSITION Entries body + +load --table Entries +[ +{"body": "Mroonga is a ＭｙＳＱＬ storage engine based on Groonga. <b>Rroonga</b> is a Ruby binding of Groonga."} +] + +select Entries --output_columns \ + --match_columns body --query 'droonga' \ + --output_columns 'highlight_html(body)' --command_version 2 Added: test/command/suite/select/function/highlight_html/no_query.expected (+37 -0) 100644 =================================================================== --- /dev/null +++ test/command/suite/select/function/highlight_html/no_query.expected 2014-08-10 06:56:57 +0900 (bbdeffa) @@ -0,0 +1,37 @@ +table_create Entries TABLE_NO_KEY +[[0,0.0,0.0],true] +column_create Entries body COLUMN_SCALAR ShortText +[[0,0.0,0.0],true] +table_create Terms TABLE_PAT_KEY ShortText --default_tokenizer TokenBigram --normalizer NormalizerAuto +[[0,0.0,0.0],true] +column_create Terms document_index COLUMN_INDEX|WITH_POSITION Entries body +[[0,0.0,0.0],true] +load --table Entries +[ +{"body": "Mroonga is a ＭｙＳＱＬ storage engine based on Groonga. <b>Rroonga</b> is a Ruby binding of Groonga."} +] +[[0,0.0,0.0],1] +select Entries --output_columns 'highlight_html(body)' --command_version 2 +[ + [ + 0, + 0.0, + 0.0 + ], + [ + [ + [ + 1 + ], + [ + [ + "highlight_html", + "null" + ] + ], + [ + "Mroonga is a ＭｙＳＱＬ storage engine based on Groonga. <b>Rroonga</b> is a Ruby binding of Groonga." + ] + ] + ] +] Added: test/command/suite/select/function/highlight_html/no_query.test (+13 -0) 100644 =================================================================== --- /dev/null +++ test/command/suite/select/function/highlight_html/no_query.test 2014-08-10 06:56:57 +0900 (f45ebb0) @@ -0,0 +1,13 @@ +table_create Entries TABLE_NO_KEY +column_create Entries body COLUMN_SCALAR ShortText + +table_create Terms TABLE_PAT_KEY ShortText --default_tokenizer TokenBigram --normalizer NormalizerAuto +column_create Terms document_index COLUMN_INDEX|WITH_POSITION Entries body + +load --table Entries +[ +{"body": "Mroonga is a ＭｙＳＱＬ storage engine based on Groonga. <b>Rroonga</b> is a Ruby binding of Groonga."} +] + +select Entries \ + --output_columns 'highlight_html(body)' --command_version 2 Added: test/command/suite/select/function/highlight_html/one_hit.expected (+37 -0) 100644 =================================================================== --- /dev/null +++ test/command/suite/select/function/highlight_html/one_hit.expected 2014-08-10 06:56:57 +0900 (969a35d) @@ -0,0 +1,37 @@ +table_create Entries TABLE_NO_KEY +[[0,0.0,0.0],true] +column_create Entries body COLUMN_SCALAR ShortText +[[0,0.0,0.0],true] +table_create Terms TABLE_PAT_KEY ShortText --default_tokenizer TokenBigram --normalizer NormalizerAuto +[[0,0.0,0.0],true] +column_create Terms document_index COLUMN_INDEX|WITH_POSITION Entries body +[[0,0.0,0.0],true] +load --table Entries +[ +{"body": "Mroonga is a ＭｙＳＱＬ storage engine based on Groonga. <b>Rroonga</b> is a Ruby binding of Groonga."} +] +[[0,0.0,0.0],1] +select Entries --output_columns --match_columns body --query 'rroonga' --output_columns 'highlight_html(body)' --command_version 2 +[ + [ + 0, + 0.0, + 0.0 + ], + [ + [ + [ + 1 + ], + [ + [ + "highlight_html", + "null" + ] + ], + [ + "Mroonga is a ＭｙＳＱＬ storage engine based on Groonga. <b><span class=\"keyword\">Rroonga</span></b> is a Ruby binding of Groonga." + ] + ] + ] +] Added: test/command/suite/select/function/highlight_html/one_hit.test (+14 -0) 100644 =================================================================== --- /dev/null +++ test/command/suite/select/function/highlight_html/one_hit.test 2014-08-10 06:56:57 +0900 (7c04975) @@ -0,0 +1,14 @@ +table_create Entries TABLE_NO_KEY +column_create Entries body COLUMN_SCALAR ShortText + +table_create Terms TABLE_PAT_KEY ShortText --default_tokenizer TokenBigram --normalizer NormalizerAuto +column_create Terms document_index COLUMN_INDEX|WITH_POSITION Entries body + +load --table Entries +[ +{"body": "Mroonga is a ＭｙＳＱＬ storage engine based on Groonga. <b>Rroonga</b> is a Ruby binding of Groonga."} +] + +select Entries --output_columns \ + --match_columns body --query 'rroonga' \ + --output_columns 'highlight_html(body)' --command_version 2 Added: test/command/suite/select/function/highlight_html/one_keyword.expected (+37 -0) 100644 =================================================================== --- /dev/null +++ test/command/suite/select/function/highlight_html/one_keyword.expected 2014-08-10 06:56:57 +0900 (4501b81) @@ -0,0 +1,37 @@ +table_create Entries TABLE_NO_KEY +[[0,0.0,0.0],true] +column_create Entries body COLUMN_SCALAR ShortText +[[0,0.0,0.0],true] +table_create Terms TABLE_PAT_KEY ShortText --default_tokenizer TokenBigram --normalizer NormalizerAuto +[[0,0.0,0.0],true] +column_create Terms document_index COLUMN_INDEX|WITH_POSITION Entries body +[[0,0.0,0.0],true] +load --table Entries +[ +{"body": "Mroonga is a ＭｙＳＱＬ storage engine based on Groonga. <b>Rroonga</b> is a Ruby binding of Groonga."} +] +[[0,0.0,0.0],1] +select Entries --output_columns --match_columns body --query 'groonga' --output_columns 'highlight_html(body)' --command_version 2 +[ + [ + 0, + 0.0, + 0.0 + ], + [ + [ + [ + 1 + ], + [ + [ + "highlight_html", + "null" + ] + ], + [ + "Mroonga is a ＭｙＳＱＬ storage engine based on <span class=\"keyword\">Groonga</span>. <b>Rroonga</b> is a Ruby binding of <span class=\"keyword\">Groonga</span>." + ] + ] + ] +] Added: test/command/suite/select/function/highlight_html/one_keyword.test (+14 -0) 100644 =================================================================== --- /dev/null +++ test/command/suite/select/function/highlight_html/one_keyword.test 2014-08-10 06:56:57 +0900 (8248cf2) @@ -0,0 +1,14 @@ +table_create Entries TABLE_NO_KEY +column_create Entries body COLUMN_SCALAR ShortText + +table_create Terms TABLE_PAT_KEY ShortText --default_tokenizer TokenBigram --normalizer NormalizerAuto +column_create Terms document_index COLUMN_INDEX|WITH_POSITION Entries body + +load --table Entries +[ +{"body": "Mroonga is a ＭｙＳＱＬ storage engine based on Groonga. <b>Rroonga</b> is a Ruby binding of Groonga."} +] + +select Entries --output_columns \ + --match_columns body --query 'groonga' \ + --output_columns 'highlight_html(body)' --command_version 2 Added: test/command/suite/select/function/highlight_html/twice_hits.expected (+37 -0) 100644 =================================================================== --- /dev/null +++ test/command/suite/select/function/highlight_html/twice_hits.expected 2014-08-10 06:56:57 +0900 (e70ed39) @@ -0,0 +1,37 @@ +table_create Entries TABLE_NO_KEY +[[0,0.0,0.0],true] +column_create Entries body COLUMN_SCALAR ShortText +[[0,0.0,0.0],true] +table_create Terms TABLE_PAT_KEY ShortText --default_tokenizer TokenBigram --normalizer NormalizerAuto +[[0,0.0,0.0],true] +column_create Terms document_index COLUMN_INDEX|WITH_POSITION Entries body +[[0,0.0,0.0],true] +load --table Entries +[ +{"body": "Mroonga is a ＭｙＳＱＬ storage engine based on Groonga. <b>Rroonga</b> is a Ruby binding of Groonga."} +] +[[0,0.0,0.0],1] +select Entries --output_columns --match_columns body --query 'Groonga' --output_columns 'highlight_html(body)' --command_version 2 +[ + [ + 0, + 0.0, + 0.0 + ], + [ + [ + [ + 1 + ], + [ + [ + "highlight_html", + "null" + ] + ], + [ + "Mroonga is a ＭｙＳＱＬ storage engine based on <span class=\"keyword\">Groonga</span>. <b>Rroonga</b> is a Ruby binding of <span class=\"keyword\">Groonga</span>." + ] + ] + ] +] Added: test/command/suite/select/function/highlight_html/twice_hits.test (+14 -0) 100644 =================================================================== --- /dev/null +++ test/command/suite/select/function/highlight_html/twice_hits.test 2014-08-10 06:56:57 +0900 (08b8637) @@ -0,0 +1,14 @@ +table_create Entries TABLE_NO_KEY +column_create Entries body COLUMN_SCALAR ShortText + +table_create Terms TABLE_PAT_KEY ShortText --default_tokenizer TokenBigram --normalizer NormalizerAuto +column_create Terms document_index COLUMN_INDEX|WITH_POSITION Entries body + +load --table Entries +[ +{"body": "Mroonga is a ＭｙＳＱＬ storage engine based on Groonga. <b>Rroonga</b> is a Ruby binding of Groonga."} +] + +select Entries --output_columns \ + --match_columns body --query 'Groonga' \ + --output_columns 'highlight_html(body)' --command_version 2 Added: test/command/suite/select/function/highlight_html/two_keywords.expected (+37 -0) 100644 =================================================================== --- /dev/null +++ test/command/suite/select/function/highlight_html/two_keywords.expected 2014-08-10 06:56:57 +0900 (27655e0) @@ -0,0 +1,37 @@ +table_create Entries TABLE_NO_KEY +[[0,0.0,0.0],true] +column_create Entries body COLUMN_SCALAR ShortText +[[0,0.0,0.0],true] +table_create Terms TABLE_PAT_KEY ShortText --default_tokenizer TokenBigram --normalizer NormalizerAuto +[[0,0.0,0.0],true] +column_create Terms document_index COLUMN_INDEX|WITH_POSITION Entries body +[[0,0.0,0.0],true] +load --table Entries +[ +{"body": "Mroonga is a ＭｙＳＱＬ storage engine based on Groonga. <b>Rroonga</b> is a Ruby binding of Groonga."} +] +[[0,0.0,0.0],1] +select Entries --output_columns --match_columns body --query 'groonga mysql' --output_columns 'highlight_html(body)' --command_version 2 +[ + [ + 0, + 0.0, + 0.0 + ], + [ + [ + [ + 1 + ], + [ + [ + "highlight_html", + "null" + ] + ], + [ + "Mroonga is a <span class=\"keyword\">ＭｙＳＱＬ</span> storage engine based on <span class=\"keyword\">Groonga</span>. <b>Rroonga</b> is a Ruby binding of <span class=\"keyword\">Groonga</span>." + ] + ] + ] +] Added: test/command/suite/select/function/highlight_html/two_keywords.test (+14 -0) 100644 =================================================================== --- /dev/null +++ test/command/suite/select/function/highlight_html/two_keywords.test 2014-08-10 06:56:57 +0900 (462f7bd) @@ -0,0 +1,14 @@ +table_create Entries TABLE_NO_KEY +column_create Entries body COLUMN_SCALAR ShortText + +table_create Terms TABLE_PAT_KEY ShortText --default_tokenizer TokenBigram --normalizer NormalizerAuto +column_create Terms document_index COLUMN_INDEX|WITH_POSITION Entries body + +load --table Entries +[ +{"body": "Mroonga is a ＭｙＳＱＬ storage engine based on Groonga. <b>Rroonga</b> is a Ruby binding of Groonga."} +] + +select Entries --output_columns \ + --match_columns body --query 'groonga mysql' \ + --output_columns 'highlight_html(body)' --command_version 2 -------------- next part -------------- HTML����������������������������...Download