diff options
| author | Vsevolod Stakhov <vsevolod@rambler-co.ru> | 2012-10-04 22:14:10 +0400 |
|---|---|---|
| committer | Vsevolod Stakhov <vsevolod@rambler-co.ru> | 2012-10-04 22:14:10 +0400 |
| commit | 3789849b7b2e617d0a287fe77490b6643f3a6b74 (patch) | |
| tree | 8415305aee39a3aad6adbccbc1941a62f3a41cf8 /src/classifiers | |
| parent | 14e1129068d55bc8de0618832d4f7d33bb1b0f06 (diff) | |
| download | rspamd-3789849b7b2e617d0a287fe77490b6643f3a6b74.tar.gz rspamd-3789849b7b2e617d0a287fe77490b6643f3a6b74.zip | |
* Add meta-classification example.
Many changes to advanced statistic and meta-classification logic.
Add example of complex meta-classification.
Diffstat (limited to 'src/classifiers')
| -rw-r--r-- | src/classifiers/bayes.c | 11 |
1 files changed, 7 insertions, 4 deletions
diff --git a/src/classifiers/bayes.c b/src/classifiers/bayes.c index cad963c4b..a80bbe0ba 100644 --- a/src/classifiers/bayes.c +++ b/src/classifiers/bayes.c @@ -391,6 +391,7 @@ bayes_learn_spam (struct classifier_ctx* ctx, statfile_pool_t *pool, struct statfile *st; stat_file_t *file; GList *cur; + gboolean skip_labels; g_assert (pool != NULL); g_assert (ctx != NULL); @@ -411,11 +412,14 @@ bayes_learn_spam (struct classifier_ctx* ctx, statfile_pool_t *pool, } } - cur = call_classifier_pre_callbacks (ctx->cfg, task, FALSE, FALSE, L); + cur = call_classifier_pre_callbacks (ctx->cfg, task, TRUE, is_spam, L); if (cur) { + skip_labels = FALSE; memory_pool_add_destructor (task->task_pool, (pool_destruct_func)g_list_free, cur); } else { + /* Do not try to learn specific statfiles if pre callback returned nil */ + skip_labels = TRUE; cur = ctx->cfg->statfiles; } @@ -435,7 +439,7 @@ bayes_learn_spam (struct classifier_ctx* ctx, statfile_pool_t *pool, while (cur) { /* Select statfiles to learn */ st = cur->data; - if (st->is_spam != is_spam) { + if (st->is_spam != is_spam || (skip_labels && st->label)) { cur = g_list_next (cur); continue; } @@ -460,8 +464,6 @@ bayes_learn_spam (struct classifier_ctx* ctx, statfile_pool_t *pool, msg_err ("cannot open statfile %s after creation", st->path); return FALSE; } - cur = g_list_next (cur); - continue; } } data.file = file; @@ -470,6 +472,7 @@ bayes_learn_spam (struct classifier_ctx* ctx, statfile_pool_t *pool, statfile_inc_revision (file); statfile_pool_unlock_file (pool, data.file); maybe_write_binlog (ctx->cfg, st, file, input); + msg_info ("increase revision for %s", st->path); cur = g_list_next (cur); } |