X-Git-Url: https://git.notmuchmail.org/git?p=notmuch;a=blobdiff_plain;f=lib%2Fdatabase.cc;h=f395061e3a73f91b3324cc97063e7ba2c149579f;hp=9cd46d4fcb7ca3c85b6d7be8340cc6b60a1a2c56;hb=8fb16e277e4d6c32bafa79ae7967e1e6ba9258e0;hpb=e9bb90ba2c18ac6c36168ccb20099ef395d7dfb6 diff --git a/lib/database.cc b/lib/database.cc index 9cd46d4f..f395061e 100644 --- a/lib/database.cc +++ b/lib/database.cc @@ -19,14 +19,19 @@ */ #include "database-private.h" +#include "parse-time-vrp.h" #include #include +#include #include -#include +#include #include /* g_free, GPtrArray, GHashTable */ +#include /* g_type_init */ + +#include /* g_mime_init */ using namespace std; @@ -44,7 +49,8 @@ typedef struct { /* Here's the current schema for our database (for NOTMUCH_DATABASE_VERSION): * - * We currently have two different types of documents: mail and directory. + * We currently have two different types of documents (mail and + * directory) and also some metadata. * * Mail document * ------------- @@ -56,8 +62,12 @@ typedef struct { * * type: mail * - * id: Unique ID of mail, (from Message-ID header or generated - * as "notmuch-sha1-. + * id: Unique ID of mail. This is from the Message-ID header + * if present and not too long (see NOTMUCH_MESSAGE_ID_MAX). + * If it's present and too long, then we use + * "notmuch-sha1-". + * If this header is not present, we use + * "notmuch-sha1-". * * thread: The ID of the thread to which the mail belongs * @@ -65,7 +75,7 @@ typedef struct { * * Multiple terms of given prefix: * - * reference: All message IDs from In-Reply-To and Re ferences + * reference: All message IDs from In-Reply-To and References * headers in the message. * * tag: Any tags associated with this message by the user. @@ -76,17 +86,22 @@ typedef struct { * STRING is the name of a file within that * directory for this mail message. * - * A mail document also has two values: + * A mail document also has four values: * * TIMESTAMP: The time_t value corresponding to the message's * Date header. * * MESSAGE_ID: The unique ID of the mail mess (see "id" above) * + * FROM: The value of the "From" header + * + * SUBJECT: The value of the "Subject" header + * * In addition, terms from the content of the message are added with * "from", "to", "attachment", and "subject" prefixes for use by the - * user in searching. But the database doesn't really care itself - * about any of these. + * user in searching. Similarly, terms from the path of the mail + * message are added with a "folder" prefix. But the database doesn't + * really care itself about any of these. * * The data portion of a mail document is empty. * @@ -118,6 +133,51 @@ typedef struct { * * The data portion of a directory document contains the path of the * directory (relative to the database path). + * + * Database metadata + * ----------------- + * Xapian allows us to store arbitrary name-value pairs as + * "metadata". We currently use the following metadata names with the + * given meanings: + * + * version The database schema version, (which is distinct + * from both the notmuch package version (see + * notmuch --version) and the libnotmuch library + * version. The version is stored as an base-10 + * ASCII integer. The initial database version + * was 1, (though a schema existed before that + * were no "version" database value existed at + * all). Successive versions are allocated as + * changes are made to the database (such as by + * indexing new fields). + * + * last_thread_id The last thread ID generated. This is stored + * as a 16-byte hexadecimal ASCII representation + * of a 64-bit unsigned integer. The first ID + * generated is 1 and the value will be + * incremented for each thread ID. + * + * thread_id_* A pre-allocated thread ID for a particular + * message. This is actually an arbitrarily large + * family of metadata name. Any particular name is + * formed by concatenating "thread_id_" with a message + * ID (or the SHA1 sum of a message ID if it is very + * long---see description of 'id' in the mail + * document). The value stored is a thread ID. + * + * These thread ID metadata values are stored + * whenever a message references a parent message + * that does not yet exist in the database. A + * thread ID will be allocated and stored, and if + * the message is later added, the stored thread + * ID will be used (and the metadata value will + * be cleared). + * + * Even before a message is added, it's + * pre-allocated thread ID is useful so that all + * descendant messages that reference this common + * parent can be recognized as belonging to the + * same thread. */ /* With these prefix values we follow the conventions published here: @@ -135,7 +195,7 @@ typedef struct { * nearly universal to all mail messages). */ -prefix_t BOOLEAN_PREFIX_INTERNAL[] = { +static prefix_t BOOLEAN_PREFIX_INTERNAL[] = { { "type", "T" }, { "reference", "XREFERENCE" }, { "replyto", "XREPLYTO" }, @@ -144,35 +204,21 @@ prefix_t BOOLEAN_PREFIX_INTERNAL[] = { { "directory-direntry", "XDDIRENTRY" }, }; -prefix_t BOOLEAN_PREFIX_EXTERNAL[] = { +static prefix_t BOOLEAN_PREFIX_EXTERNAL[] = { { "thread", "G" }, { "tag", "K" }, { "is", "K" }, { "id", "Q" } }; -prefix_t PROBABILISTIC_PREFIX[]= { +static prefix_t PROBABILISTIC_PREFIX[]= { { "from", "XFROM" }, { "to", "XTO" }, { "attachment", "XATTACHMENT" }, - { "subject", "XSUBJECT"} + { "subject", "XSUBJECT"}, + { "folder", "XFOLDER"} }; -int -_internal_error (const char *format, ...) -{ - va_list va_args; - - va_start (va_args, format); - - fprintf (stderr, "Internal error: "); - vfprintf (stderr, format, va_args); - - exit (1); - - return 1; -} - const char * _find_prefix (const char *name) { @@ -222,6 +268,10 @@ notmuch_status_to_string (notmuch_status_t status) return "Tag value is too long (exceeds NOTMUCH_TAG_MAX)"; case NOTMUCH_STATUS_UNBALANCED_FREEZE_THAW: return "Unbalanced number of calls to notmuch_message_freeze/thaw"; + case NOTMUCH_STATUS_UNBALANCED_ATOMIC: + return "Unbalanced number of calls to notmuch_database_begin_atomic/end_atomic"; + case NOTMUCH_STATUS_UNSUPPORTED_OPERATION: + return "Unsupported operation"; default: case NOTMUCH_STATUS_LAST_STATUS: return "Unknown error status value"; @@ -290,20 +340,58 @@ find_document_for_doc_id (notmuch_database_t *notmuch, unsigned doc_id) return notmuch->xapian_db->get_document (doc_id); } -notmuch_message_t * +/* Generate a compressed version of 'message_id' of the form: + * + * notmuch-sha1- + */ +static char * +_message_id_compressed (void *ctx, const char *message_id) +{ + char *sha1, *compressed; + + sha1 = notmuch_sha1_of_string (message_id); + + compressed = talloc_asprintf (ctx, "notmuch-sha1-%s", sha1); + free (sha1); + + return compressed; +} + +notmuch_status_t notmuch_database_find_message (notmuch_database_t *notmuch, - const char *message_id) + const char *message_id, + notmuch_message_t **message_ret) { notmuch_private_status_t status; unsigned int doc_id; - status = _notmuch_database_find_unique_doc_id (notmuch, "id", - message_id, &doc_id); + if (message_ret == NULL) + return NOTMUCH_STATUS_NULL_POINTER; - if (status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND) - return NULL; + if (strlen (message_id) > NOTMUCH_MESSAGE_ID_MAX) + message_id = _message_id_compressed (notmuch, message_id); + + try { + status = _notmuch_database_find_unique_doc_id (notmuch, "id", + message_id, &doc_id); + + if (status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND) + *message_ret = NULL; + else { + *message_ret = _notmuch_message_create (notmuch, notmuch, doc_id, + NULL); + if (*message_ret == NULL) + return NOTMUCH_STATUS_OUT_OF_MEMORY; + } - return _notmuch_message_create (notmuch, notmuch, doc_id, NULL); + return NOTMUCH_STATUS_SUCCESS; + } catch (const Xapian::Error &error) { + fprintf (stderr, "A Xapian exception occurred finding message: %s.\n", + error.get_msg().c_str()); + notmuch->exception_reported = TRUE; + *message_ret = NULL; + return NOTMUCH_STATUS_XAPIAN_EXCEPTION; + } } /* Advance 'str' past any whitespace or RFC 822 comments. A comment is @@ -344,7 +432,7 @@ skip_space_and_comments (const char **str) } /* Parse an RFC 822 message-id, discarding whitespace, any RFC 822 - * comments, and the '<' and '>' delimeters. + * comments, and the '<' and '>' delimiters. * * If not NULL, then *next will be made to point to the first character * not parsed, (possibly pointing to the final '\0' terminator. @@ -417,8 +505,10 @@ _parse_message_id (void *ctx, const char *message_id, const char **next) * 'message_id' in the result (to avoid mass confusion when a single * message references itself cyclically---and yes, mail messages are * not infrequent in the wild that do this---don't ask me why). -*/ -static void + * + * Return the last reference parsed, if it is not equal to message_id. + */ +static char * parse_references (void *ctx, const char *message_id, GHashTable *hash, @@ -427,7 +517,7 @@ parse_references (void *ctx, char *ref; if (refs == NULL || *refs == '\0') - return; + return NULL; while (*refs) { ref = _parse_message_id (ctx, refs, &refs); @@ -435,11 +525,23 @@ parse_references (void *ctx, if (ref && strcmp (ref, message_id)) g_hash_table_insert (hash, ref, NULL); } + + /* The return value of this function is used to add a parent + * reference to the database. We should avoid making a message + * its own parent, thus the following check. + */ + + if (ref && strcmp(ref, message_id)) { + return ref; + } else { + return NULL; + } } -notmuch_database_t * -notmuch_database_create (const char *path) +notmuch_status_t +notmuch_database_create (const char *path, notmuch_database_t **database) { + notmuch_status_t status = NOTMUCH_STATUS_SUCCESS; notmuch_database_t *notmuch = NULL; char *notmuch_path = NULL; struct stat st; @@ -447,6 +549,7 @@ notmuch_database_create (const char *path) if (path == NULL) { fprintf (stderr, "Error: Cannot create a database for a NULL path.\n"); + status = NOTMUCH_STATUS_NULL_POINTER; goto DONE; } @@ -454,12 +557,14 @@ notmuch_database_create (const char *path) if (err) { fprintf (stderr, "Error: Cannot create database at %s: %s.\n", path, strerror (errno)); + status = NOTMUCH_STATUS_FILE_ERROR; goto DONE; } if (! S_ISDIR (st.st_mode)) { fprintf (stderr, "Error: Cannot create database at %s: Not a directory.\n", path); + status = NOTMUCH_STATUS_FILE_ERROR; goto DONE; } @@ -470,18 +575,30 @@ notmuch_database_create (const char *path) if (err) { fprintf (stderr, "Error: Cannot create directory %s: %s.\n", notmuch_path, strerror (errno)); + status = NOTMUCH_STATUS_FILE_ERROR; goto DONE; } - notmuch = notmuch_database_open (path, - NOTMUCH_DATABASE_MODE_READ_WRITE); - notmuch_database_upgrade (notmuch, NULL, NULL); + status = notmuch_database_open (path, + NOTMUCH_DATABASE_MODE_READ_WRITE, + ¬much); + if (status) + goto DONE; + status = notmuch_database_upgrade (notmuch, NULL, NULL); + if (status) { + notmuch_database_close(notmuch); + notmuch = NULL; + } DONE: if (notmuch_path) talloc_free (notmuch_path); - return notmuch; + if (database) + *database = notmuch; + else + talloc_free (notmuch); + return status; } notmuch_status_t @@ -495,19 +612,29 @@ _notmuch_database_ensure_writable (notmuch_database_t *notmuch) return NOTMUCH_STATUS_SUCCESS; } -notmuch_database_t * +notmuch_status_t notmuch_database_open (const char *path, - notmuch_database_mode_t mode) + notmuch_database_mode_t mode, + notmuch_database_t **database) { + notmuch_status_t status = NOTMUCH_STATUS_SUCCESS; + void *local = talloc_new (NULL); notmuch_database_t *notmuch = NULL; - char *notmuch_path = NULL, *xapian_path = NULL; + char *notmuch_path, *xapian_path; struct stat st; int err; unsigned int i, version; + static int initialized = 0; - if (asprintf (¬much_path, "%s/%s", path, ".notmuch") == -1) { - notmuch_path = NULL; + if (path == NULL) { + fprintf (stderr, "Error: Cannot open a database for a NULL path.\n"); + status = NOTMUCH_STATUS_NULL_POINTER; + goto DONE; + } + + if (! (notmuch_path = talloc_asprintf (local, "%s/%s", path, ".notmuch"))) { fprintf (stderr, "Out of memory\n"); + status = NOTMUCH_STATUS_OUT_OF_MEMORY; goto DONE; } @@ -515,16 +642,28 @@ notmuch_database_open (const char *path, if (err) { fprintf (stderr, "Error opening database at %s: %s\n", notmuch_path, strerror (errno)); + status = NOTMUCH_STATUS_FILE_ERROR; goto DONE; } - if (asprintf (&xapian_path, "%s/%s", notmuch_path, "xapian") == -1) { - xapian_path = NULL; + if (! (xapian_path = talloc_asprintf (local, "%s/%s", notmuch_path, "xapian"))) { fprintf (stderr, "Out of memory\n"); + status = NOTMUCH_STATUS_OUT_OF_MEMORY; goto DONE; } - notmuch = talloc (NULL, notmuch_database_t); + /* Initialize the GLib type system and threads */ +#if !GLIB_CHECK_VERSION(2, 35, 1) + g_type_init (); +#endif + + /* Initialize gmime */ + if (! initialized) { + g_mime_init (GMIME_ENABLE_RFC2047_WORKAROUNDS); + initialized = 1; + } + + notmuch = talloc_zero (NULL, notmuch_database_t); notmuch->exception_reported = FALSE; notmuch->path = talloc_strdup (notmuch, path); @@ -533,6 +672,7 @@ notmuch_database_open (const char *path, notmuch->needs_upgrade = FALSE; notmuch->mode = mode; + notmuch->atomic_nesting = 0; try { string last_thread_id; @@ -549,8 +689,9 @@ notmuch_database_open (const char *path, " read-write mode.\n", notmuch_path, version, NOTMUCH_DATABASE_VERSION); notmuch->mode = NOTMUCH_DATABASE_MODE_READ_ONLY; - notmuch_database_close (notmuch); + notmuch_database_destroy (notmuch); notmuch = NULL; + status = NOTMUCH_STATUS_FILE_ERROR; goto DONE; } @@ -571,6 +712,7 @@ notmuch_database_open (const char *path, } } + notmuch->last_doc_id = notmuch->xapian_db->get_lastdocid (); last_thread_id = notmuch->xapian_db->get_metadata ("last_thread_id"); if (last_thread_id.empty ()) { notmuch->last_thread_id = 0; @@ -588,12 +730,14 @@ notmuch_database_open (const char *path, notmuch->term_gen = new Xapian::TermGenerator; notmuch->term_gen->set_stemmer (Xapian::Stem ("english")); notmuch->value_range_processor = new Xapian::NumberValueRangeProcessor (NOTMUCH_VALUE_TIMESTAMP); + notmuch->date_range_processor = new ParseTimeValueRangeProcessor (NOTMUCH_VALUE_TIMESTAMP); notmuch->query_parser->set_default_op (Xapian::Query::OP_AND); notmuch->query_parser->set_database (*notmuch->xapian_db); notmuch->query_parser->set_stemmer (Xapian::Stem ("english")); notmuch->query_parser->set_stemming_strategy (Xapian::QueryParser::STEM_SOME); notmuch->query_parser->add_valuerangeprocessor (notmuch->value_range_processor); + notmuch->query_parser->add_valuerangeprocessor (notmuch->date_range_processor); for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_EXTERNAL); i++) { prefix_t *prefix = &BOOLEAN_PREFIX_EXTERNAL[i]; @@ -608,23 +752,27 @@ notmuch_database_open (const char *path, } catch (const Xapian::Error &error) { fprintf (stderr, "A Xapian exception occurred opening database: %s\n", error.get_msg().c_str()); + notmuch_database_destroy (notmuch); notmuch = NULL; + status = NOTMUCH_STATUS_XAPIAN_EXCEPTION; } DONE: - if (notmuch_path) - free (notmuch_path); - if (xapian_path) - free (xapian_path); + talloc_free (local); - return notmuch; + if (database) + *database = notmuch; + else + talloc_free (notmuch); + return status; } void notmuch_database_close (notmuch_database_t *notmuch) { try { - if (notmuch->mode == NOTMUCH_DATABASE_MODE_READ_WRITE) + if (notmuch->xapian_db != NULL && + notmuch->mode == NOTMUCH_DATABASE_MODE_READ_WRITE) (static_cast (notmuch->xapian_db))->flush (); } catch (const Xapian::Error &error) { if (! notmuch->exception_reported) { @@ -633,10 +781,213 @@ notmuch_database_close (notmuch_database_t *notmuch) } } + /* Many Xapian objects (and thus notmuch objects) hold references to + * the database, so merely deleting the database may not suffice to + * close it. Thus, we explicitly close it here. */ + if (notmuch->xapian_db != NULL) { + try { + notmuch->xapian_db->close(); + } catch (const Xapian::Error &error) { + /* do nothing */ + } + } + delete notmuch->term_gen; + notmuch->term_gen = NULL; delete notmuch->query_parser; + notmuch->query_parser = NULL; delete notmuch->xapian_db; + notmuch->xapian_db = NULL; delete notmuch->value_range_processor; + notmuch->value_range_processor = NULL; + delete notmuch->date_range_processor; + notmuch->date_range_processor = NULL; +} + +#if HAVE_XAPIAN_COMPACT +static int +unlink_cb (const char *path, + unused (const struct stat *sb), + unused (int type), + unused (struct FTW *ftw)) +{ + return remove (path); +} + +static int +rmtree (const char *path) +{ + return nftw (path, unlink_cb, 64, FTW_DEPTH | FTW_PHYS); +} + +class NotmuchCompactor : public Xapian::Compactor +{ + notmuch_compact_status_cb_t status_cb; + void *status_closure; + +public: + NotmuchCompactor(notmuch_compact_status_cb_t cb, void *closure) : + status_cb (cb), status_closure (closure) { } + + virtual void + set_status (const std::string &table, const std::string &status) + { + char *msg; + + if (status_cb == NULL) + return; + + if (status.length () == 0) + msg = talloc_asprintf (NULL, "compacting table %s", table.c_str()); + else + msg = talloc_asprintf (NULL, " %s", status.c_str()); + + if (msg == NULL) { + return; + } + + status_cb (msg, status_closure); + talloc_free (msg); + } +}; + +/* Compacts the given database, optionally saving the original database + * in backup_path. Additionally, a callback function can be provided to + * give the user feedback on the progress of the (likely long-lived) + * compaction process. + * + * The backup path must point to a directory on the same volume as the + * original database. Passing a NULL backup_path will result in the + * uncompacted database being deleted after compaction has finished. + * Note that the database write lock will be held during the + * compaction process to protect data integrity. + */ +notmuch_status_t +notmuch_database_compact (const char *path, + const char *backup_path, + notmuch_compact_status_cb_t status_cb, + void *closure) +{ + void *local; + char *notmuch_path, *xapian_path, *compact_xapian_path; + notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS; + notmuch_database_t *notmuch = NULL; + struct stat statbuf; + notmuch_bool_t keep_backup; + + local = talloc_new (NULL); + if (! local) + return NOTMUCH_STATUS_OUT_OF_MEMORY; + + ret = notmuch_database_open (path, NOTMUCH_DATABASE_MODE_READ_WRITE, ¬much); + if (ret) { + goto DONE; + } + + if (! (notmuch_path = talloc_asprintf (local, "%s/%s", path, ".notmuch"))) { + ret = NOTMUCH_STATUS_OUT_OF_MEMORY; + goto DONE; + } + + if (! (xapian_path = talloc_asprintf (local, "%s/%s", notmuch_path, "xapian"))) { + ret = NOTMUCH_STATUS_OUT_OF_MEMORY; + goto DONE; + } + + if (! (compact_xapian_path = talloc_asprintf (local, "%s.compact", xapian_path))) { + ret = NOTMUCH_STATUS_OUT_OF_MEMORY; + goto DONE; + } + + if (backup_path == NULL) { + if (! (backup_path = talloc_asprintf (local, "%s.old", xapian_path))) { + ret = NOTMUCH_STATUS_OUT_OF_MEMORY; + goto DONE; + } + keep_backup = FALSE; + } + else { + keep_backup = TRUE; + } + + if (stat (backup_path, &statbuf) != -1) { + fprintf (stderr, "Path already exists: %s\n", backup_path); + ret = NOTMUCH_STATUS_FILE_ERROR; + goto DONE; + } + if (errno != ENOENT) { + fprintf (stderr, "Unknown error while stat()ing path: %s\n", + strerror (errno)); + ret = NOTMUCH_STATUS_FILE_ERROR; + goto DONE; + } + + /* Unconditionally attempt to remove old work-in-progress database (if + * any). This is "protected" by database lock. If this fails due to write + * errors (etc), the following code will fail and provide error message. + */ + (void) rmtree (compact_xapian_path); + + try { + NotmuchCompactor compactor (status_cb, closure); + + compactor.set_renumber (false); + compactor.add_source (xapian_path); + compactor.set_destdir (compact_xapian_path); + compactor.compact (); + } catch (const Xapian::Error &error) { + fprintf (stderr, "Error while compacting: %s\n", error.get_msg().c_str()); + ret = NOTMUCH_STATUS_XAPIAN_EXCEPTION; + goto DONE; + } + + if (rename (xapian_path, backup_path)) { + fprintf (stderr, "Error moving %s to %s: %s\n", + xapian_path, backup_path, strerror (errno)); + ret = NOTMUCH_STATUS_FILE_ERROR; + goto DONE; + } + + if (rename (compact_xapian_path, xapian_path)) { + fprintf (stderr, "Error moving %s to %s: %s\n", + compact_xapian_path, xapian_path, strerror (errno)); + ret = NOTMUCH_STATUS_FILE_ERROR; + goto DONE; + } + + if (! keep_backup) { + if (rmtree (backup_path)) { + fprintf (stderr, "Error removing old database %s: %s\n", + backup_path, strerror (errno)); + ret = NOTMUCH_STATUS_FILE_ERROR; + goto DONE; + } + } + + DONE: + if (notmuch) + notmuch_database_destroy (notmuch); + + talloc_free (local); + + return ret; +} +#else +notmuch_status_t +notmuch_database_compact (unused (const char *path), + unused (const char *backup_path), + unused (notmuch_compact_status_cb_t status_cb), + unused (void *closure)) +{ + fprintf (stderr, "notmuch was compiled against a xapian version lacking compaction support.\n"); + return NOTMUCH_STATUS_UNSUPPORTED_OPERATION; +} +#endif + +void +notmuch_database_destroy (notmuch_database_t *notmuch) +{ + notmuch_database_close (notmuch); talloc_free (notmuch); } @@ -808,8 +1159,8 @@ notmuch_database_upgrade (notmuch_database_t *notmuch, mtime = Xapian::sortable_unserialise ( document.get_value (NOTMUCH_VALUE_TIMESTAMP)); - directory = notmuch_database_get_directory (notmuch, - term.c_str() + 10); + directory = _notmuch_directory_create (notmuch, term.c_str() + 10, + NOTMUCH_FIND_CREATE, &status); notmuch_directory_set_mtime (directory, mtime); notmuch_directory_destroy (directory); } @@ -895,6 +1246,61 @@ notmuch_database_upgrade (notmuch_database_t *notmuch, return NOTMUCH_STATUS_SUCCESS; } +notmuch_status_t +notmuch_database_begin_atomic (notmuch_database_t *notmuch) +{ + if (notmuch->mode == NOTMUCH_DATABASE_MODE_READ_ONLY || + notmuch->atomic_nesting > 0) + goto DONE; + + try { + (static_cast (notmuch->xapian_db))->begin_transaction (false); + } catch (const Xapian::Error &error) { + fprintf (stderr, "A Xapian exception occurred beginning transaction: %s.\n", + error.get_msg().c_str()); + notmuch->exception_reported = TRUE; + return NOTMUCH_STATUS_XAPIAN_EXCEPTION; + } + +DONE: + notmuch->atomic_nesting++; + return NOTMUCH_STATUS_SUCCESS; +} + +notmuch_status_t +notmuch_database_end_atomic (notmuch_database_t *notmuch) +{ + Xapian::WritableDatabase *db; + + if (notmuch->atomic_nesting == 0) + return NOTMUCH_STATUS_UNBALANCED_ATOMIC; + + if (notmuch->mode == NOTMUCH_DATABASE_MODE_READ_ONLY || + notmuch->atomic_nesting > 1) + goto DONE; + + db = static_cast (notmuch->xapian_db); + try { + db->commit_transaction (); + + /* This is a hack for testing. Xapian never flushes on a + * non-flushed commit, even if the flush threshold is 1. + * However, we rely on flushing to test atomicity. */ + const char *thresh = getenv ("XAPIAN_FLUSH_THRESHOLD"); + if (thresh && atoi (thresh) == 1) + db->flush (); + } catch (const Xapian::Error &error) { + fprintf (stderr, "A Xapian exception occurred committing transaction: %s.\n", + error.get_msg().c_str()); + notmuch->exception_reported = TRUE; + return NOTMUCH_STATUS_XAPIAN_EXCEPTION; + } + +DONE: + notmuch->atomic_nesting--; + return NOTMUCH_STATUS_SUCCESS; +} + /* We allow the user to use arbitrarily long paths for directories. But * we have a term-length limit. So if we exceed that, we'll use the * SHA-1 of the path for the database term. @@ -995,9 +1401,17 @@ _notmuch_database_split_path (void *ctx, return NOTMUCH_STATUS_SUCCESS; } +/* Find the document ID of the specified directory. + * + * If (flags & NOTMUCH_FIND_CREATE), a new directory document will be + * created if one does not exist for 'path'. Otherwise, if the + * directory document does not exist, this sets *directory_id to + * ((unsigned int)-1) and returns NOTMUCH_STATUS_SUCCESS. + */ notmuch_status_t _notmuch_database_find_directory_id (notmuch_database_t *notmuch, const char *path, + notmuch_find_flags_t flags, unsigned int *directory_id) { notmuch_directory_t *directory; @@ -1008,8 +1422,8 @@ _notmuch_database_find_directory_id (notmuch_database_t *notmuch, return NOTMUCH_STATUS_SUCCESS; } - directory = _notmuch_directory_create (notmuch, path, &status); - if (status) { + directory = _notmuch_directory_create (notmuch, path, flags, &status); + if (status || !directory) { *directory_id = -1; return status; } @@ -1038,13 +1452,16 @@ _notmuch_database_get_directory_path (void *ctx, * database path), return a new string (with 'ctx' as the talloc * owner) suitable for use as a direntry term value. * - * The necessary directory documents will be created in the database - * as needed. + * If (flags & NOTMUCH_FIND_CREATE), the necessary directory documents + * will be created in the database as needed. Otherwise, if the + * necessary directory documents do not exist, this sets + * *direntry to NULL and returns NOTMUCH_STATUS_SUCCESS. */ notmuch_status_t _notmuch_database_filename_to_direntry (void *ctx, notmuch_database_t *notmuch, const char *filename, + notmuch_find_flags_t flags, char **direntry) { const char *relative, *directory, *basename; @@ -1058,10 +1475,12 @@ _notmuch_database_filename_to_direntry (void *ctx, if (status) return status; - status = _notmuch_database_find_directory_id (notmuch, directory, + status = _notmuch_database_find_directory_id (notmuch, directory, flags, &directory_id); - if (status) + if (status || directory_id == (unsigned int)-1) { + *direntry = NULL; return status; + } *direntry = talloc_asprintf (ctx, "%u:%s", directory_id, basename); @@ -1070,7 +1489,7 @@ _notmuch_database_filename_to_direntry (void *ctx, /* Given a legal 'path' for the database, return the relative path. * - * The return value will be a pointer to the originl path contents, + * The return value will be a pointer to the original path contents, * and will be either the original string (if 'path' was relative) or * a portion of the string (if path was absolute and begins with the * database path). @@ -1102,13 +1521,52 @@ _notmuch_database_relative_path (notmuch_database_t *notmuch, return relative; } -notmuch_directory_t * +notmuch_status_t notmuch_database_get_directory (notmuch_database_t *notmuch, - const char *path) + const char *path, + notmuch_directory_t **directory) { notmuch_status_t status; - return _notmuch_directory_create (notmuch, path, &status); + if (directory == NULL) + return NOTMUCH_STATUS_NULL_POINTER; + *directory = NULL; + + try { + *directory = _notmuch_directory_create (notmuch, path, + NOTMUCH_FIND_LOOKUP, &status); + } catch (const Xapian::Error &error) { + fprintf (stderr, "A Xapian exception occurred getting directory: %s.\n", + error.get_msg().c_str()); + notmuch->exception_reported = TRUE; + status = NOTMUCH_STATUS_XAPIAN_EXCEPTION; + } + return status; +} + +/* Allocate a document ID that satisfies the following criteria: + * + * 1. The ID does not exist for any document in the Xapian database + * + * 2. The ID was not previously returned from this function + * + * 3. The ID is the smallest integer satisfying (1) and (2) + * + * This function will trigger an internal error if these constraints + * cannot all be satisfied, (that is, the pool of available document + * IDs has been exhausted). + */ +unsigned int +_notmuch_database_generate_doc_id (notmuch_database_t *notmuch) +{ + assert (notmuch->last_doc_id >= notmuch->xapian_db->get_lastdocid ()); + + notmuch->last_doc_id++; + + if (notmuch->last_doc_id == 0) + INTERNAL_ERROR ("Xapian document IDs are exhausted.\n"); + + return notmuch->last_doc_id; } static const char * @@ -1133,12 +1591,18 @@ _notmuch_database_generate_thread_id (notmuch_database_t *notmuch) static char * _get_metadata_thread_id_key (void *ctx, const char *message_id) { - return talloc_asprintf (ctx, "thread_id_%s", message_id); + if (strlen (message_id) > NOTMUCH_MESSAGE_ID_MAX) + message_id = _message_id_compressed (ctx, message_id); + + return talloc_asprintf (ctx, NOTMUCH_METADATA_THREAD_ID_PREFIX "%s", + message_id); } /* Find the thread ID to which the message with 'message_id' belongs. * - * Always returns a newly talloced string belonging to 'ctx'. + * Note: 'thread_id_ret' must not be NULL! + * On success '*thread_id_ret' is set to a newly talloced string belonging to + * 'ctx'. * * Note: If there is no message in the database with the given * 'message_id' then a new thread_id will be allocated for this @@ -1146,42 +1610,53 @@ _get_metadata_thread_id_key (void *ctx, const char *message_id) * thread ID can be looked up if the message is added to the database * later). */ -static const char * +static notmuch_status_t _resolve_message_id_to_thread_id (notmuch_database_t *notmuch, void *ctx, - const char *message_id) + const char *message_id, + const char **thread_id_ret) { + notmuch_status_t status; notmuch_message_t *message; - const char *ret = NULL; - - message = notmuch_database_find_message (notmuch, message_id); - /* If we haven't seen that message yet then check if we have already - * generated a dummy id for it and stored it in the metadata. - * If not then we generate a new thread id. - * This ensures that we can thread messages even when we haven't received - * the root (yet?) - */ - if (message == NULL) { - Xapian::WritableDatabase *db = static_cast (notmuch->xapian_db); - char * metadata_key = _get_metadata_thread_id_key (ctx, message_id); - string thread_id = notmuch->xapian_db->get_metadata(metadata_key); - if (thread_id.empty()) { - ret = _notmuch_database_generate_thread_id(notmuch); - db->set_metadata(metadata_key, ret); - } else { - ret = thread_id.c_str(); - } - talloc_free (metadata_key); - goto DONE; - } - - ret = talloc_steal (ctx, notmuch_message_get_thread_id (message)); + string thread_id_string; + char *metadata_key; + Xapian::WritableDatabase *db; + + status = notmuch_database_find_message (notmuch, message_id, &message); + + if (status) + return status; + + if (message) { + *thread_id_ret = talloc_steal (ctx, + notmuch_message_get_thread_id (message)); - DONE: - if (message) notmuch_message_destroy (message); - return ret; + return NOTMUCH_STATUS_SUCCESS; + } + + /* Message has not been seen yet. + * + * We may have seen a reference to it already, in which case, we + * can return the thread ID stored in the metadata. Otherwise, we + * generate a new thread ID and store it there. + */ + db = static_cast (notmuch->xapian_db); + metadata_key = _get_metadata_thread_id_key (ctx, message_id); + thread_id_string = notmuch->xapian_db->get_metadata (metadata_key); + + if (thread_id_string.empty()) { + *thread_id_ret = talloc_strdup (ctx, + _notmuch_database_generate_thread_id (notmuch)); + db->set_metadata (metadata_key, *thread_id_ret); + } else { + *thread_id_ret = talloc_strdup (ctx, thread_id_string.c_str()); + } + + talloc_free (metadata_key); + + return NOTMUCH_STATUS_SUCCESS; } static notmuch_status_t @@ -1234,52 +1709,59 @@ _notmuch_database_link_message_to_parents (notmuch_database_t *notmuch, { GHashTable *parents = NULL; const char *refs, *in_reply_to, *in_reply_to_message_id; + const char *last_ref_message_id, *this_message_id; GList *l, *keys = NULL; notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS; parents = g_hash_table_new_full (g_str_hash, g_str_equal, _my_talloc_free_for_g_hash, NULL); + this_message_id = notmuch_message_get_message_id (message); refs = notmuch_message_file_get_header (message_file, "references"); - parse_references (message, notmuch_message_get_message_id (message), - parents, refs); + last_ref_message_id = parse_references (message, + this_message_id, + parents, refs); in_reply_to = notmuch_message_file_get_header (message_file, "in-reply-to"); - parse_references (message, notmuch_message_get_message_id (message), - parents, in_reply_to); - - /* Carefully avoid adding any self-referential in-reply-to term. */ - in_reply_to_message_id = _parse_message_id (message, in_reply_to, NULL); - if (in_reply_to_message_id && - strcmp (in_reply_to_message_id, - notmuch_message_get_message_id (message))) - { + in_reply_to_message_id = parse_references (message, + this_message_id, + parents, in_reply_to); + + /* For the parent of this message, use the last message ID of the + * References header, if available. If not, fall back to the + * first message ID in the In-Reply-To header. */ + if (last_ref_message_id) { + _notmuch_message_add_term (message, "replyto", + last_ref_message_id); + } else if (in_reply_to_message_id) { _notmuch_message_add_term (message, "replyto", - _parse_message_id (message, in_reply_to, NULL)); + in_reply_to_message_id); } keys = g_hash_table_get_keys (parents); for (l = keys; l; l = l->next) { char *parent_message_id; - const char *parent_thread_id; + const char *parent_thread_id = NULL; parent_message_id = (char *) l->data; - parent_thread_id = _resolve_message_id_to_thread_id (notmuch, - message, - parent_message_id); - if (parent_thread_id == NULL) { - _notmuch_message_add_term (message, "reference", - parent_message_id); - } else { - if (*thread_id == NULL) { - *thread_id = talloc_strdup (message, parent_thread_id); - _notmuch_message_add_term (message, "thread", *thread_id); - } else if (strcmp (*thread_id, parent_thread_id)) { - ret = _merge_threads (notmuch, *thread_id, parent_thread_id); - if (ret) - goto DONE; - } + _notmuch_message_add_term (message, "reference", + parent_message_id); + + ret = _resolve_message_id_to_thread_id (notmuch, + message, + parent_message_id, + &parent_thread_id); + if (ret) + goto DONE; + + if (*thread_id == NULL) { + *thread_id = talloc_strdup (message, parent_thread_id); + _notmuch_message_add_term (message, "thread", *thread_id); + } else if (strcmp (*thread_id, parent_thread_id)) { + ret = _merge_threads (notmuch, *thread_id, parent_thread_id); + if (ret) + goto DONE; } } @@ -1343,18 +1825,27 @@ _notmuch_database_link_message_to_children (notmuch_database_t *notmuch, /* Given a (mostly empty) 'message' and its corresponding * 'message_file' link it to existing threads in the database. * - * We first look at 'message_file' and its link-relevant headers - * (References and In-Reply-To) for message IDs. We also look in the - * database for existing message that reference 'message'. In either - * case, we will assign to the current message the first thread_id + * The first check is in the metadata of the database to see if we + * have pre-allocated a thread_id in advance for this message, (which + * would have happened if a message was previously added that + * referenced this one). + * + * Second, we look at 'message_file' and its link-relevant headers + * (References and In-Reply-To) for message IDs. + * + * Finally, we look in the database for existing message that + * reference 'message'. + * + * In all cases, we assign to the current message the first thread_id * found (through either parent or child). We will also merge any * existing, distinct threads where this message belongs to both, - * (which is not uncommon when mesages are processed out of order). + * (which is not uncommon when messages are processed out of order). * - * Finally, if not thread ID has been found through parent or child, - * we call _notmuch_message_generate_thread_id to generate a new - * generates a new thread ID if the message doesn't connect to any - * existing threads. + * Finally, if no thread ID has been found through parent or child, we + * call _notmuch_message_generate_thread_id to generate a new thread + * ID. This should only happen for new, top-level messages, (no + * References or In-Reply-To header in this message, and no previously + * added message refers to this message). */ static notmuch_status_t _notmuch_database_link_message (notmuch_database_t *notmuch, @@ -1362,17 +1853,27 @@ _notmuch_database_link_message (notmuch_database_t *notmuch, notmuch_message_file_t *message_file) { notmuch_status_t status; - const char *thread_id = NULL; - char *metadata_key = _get_metadata_thread_id_key (message, - notmuch_message_get_message_id (message)); + const char *message_id, *thread_id = NULL; + char *metadata_key; + string stored_id; + + message_id = notmuch_message_get_message_id (message); + metadata_key = _get_metadata_thread_id_key (message, message_id); + /* Check if we have already seen related messages to this one. * If we have then use the thread_id that we stored at that time. */ - string stored_id = notmuch->xapian_db->get_metadata (metadata_key); - if (!stored_id.empty()) { - Xapian::WritableDatabase *db = static_cast (notmuch->xapian_db); + stored_id = notmuch->xapian_db->get_metadata (metadata_key); + if (! stored_id.empty()) { + Xapian::WritableDatabase *db; + + db = static_cast (notmuch->xapian_db); + + /* Clear the metadata for this message ID. We don't need it + * anymore. */ db->set_metadata (metadata_key, ""); thread_id = stored_id.c_str(); + _notmuch_message_add_term (message, "thread", thread_id); } talloc_free (metadata_key); @@ -1405,7 +1906,7 @@ notmuch_database_add_message (notmuch_database_t *notmuch, { notmuch_message_file_t *message_file; notmuch_message_t *message = NULL; - notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS; + notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS, ret2; notmuch_private_status_t private_status; const char *date, *header; @@ -1423,6 +1924,12 @@ notmuch_database_add_message (notmuch_database_t *notmuch, if (message_file == NULL) return NOTMUCH_STATUS_FILE_ERROR; + /* Adding a message may change many documents. Do this all + * atomically. */ + ret = notmuch_database_begin_atomic (notmuch); + if (ret) + goto DONE; + notmuch_message_file_restrict_headers (message_file, "date", "from", @@ -1463,10 +1970,12 @@ notmuch_database_add_message (notmuch_database_t *notmuch, if (message_id == NULL) message_id = talloc_strdup (message_file, header); - /* Reject a Message ID that's too long. */ - if (message_id && strlen (message_id) + 1 > NOTMUCH_TERM_MAX) { + /* If a message ID is too long, substitute its sha1 instead. */ + if (message_id && strlen (message_id) > NOTMUCH_MESSAGE_ID_MAX) { + char *compressed = _message_id_compressed (message_file, + message_id); talloc_free (message_id); - message_id = NULL; + message_id = compressed; } } @@ -1514,9 +2023,11 @@ notmuch_database_add_message (notmuch_database_t *notmuch, goto DONE; date = notmuch_message_file_get_header (message_file, "date"); - _notmuch_message_set_date (message, date); + _notmuch_message_set_header_values (message, date, from, subject); - _notmuch_message_index_file (message, filename); + ret = _notmuch_message_index_file (message, filename); + if (ret) + goto DONE; } else { ret = NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID; } @@ -1524,7 +2035,7 @@ notmuch_database_add_message (notmuch_database_t *notmuch, _notmuch_message_sync (message); } catch (const Xapian::Error &error) { fprintf (stderr, "A Xapian exception occurred adding message: %s.\n", - error.get_description().c_str()); + error.get_msg().c_str()); notmuch->exception_reported = TRUE; ret = NOTMUCH_STATUS_XAPIAN_EXCEPTION; goto DONE; @@ -1532,7 +2043,8 @@ notmuch_database_add_message (notmuch_database_t *notmuch, DONE: if (message) { - if (ret == NOTMUCH_STATUS_SUCCESS && message_ret) + if ((ret == NOTMUCH_STATUS_SUCCESS || + ret == NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID) && message_ret) *message_ret = message; else notmuch_message_destroy (message); @@ -1541,6 +2053,12 @@ notmuch_database_add_message (notmuch_database_t *notmuch, if (message_file) notmuch_message_file_close (message_file); + ret2 = notmuch_database_end_atomic (notmuch); + if ((ret == NOTMUCH_STATUS_SUCCESS || + ret == NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID) && + ret2 != NOTMUCH_STATUS_SUCCESS) + ret = ret2; + return ret; } @@ -1548,95 +2066,119 @@ notmuch_status_t notmuch_database_remove_message (notmuch_database_t *notmuch, const char *filename) { - Xapian::WritableDatabase *db; - void *local = talloc_new (notmuch); + notmuch_status_t status; + notmuch_message_t *message; + + status = notmuch_database_find_message_by_filename (notmuch, filename, + &message); + + if (status == NOTMUCH_STATUS_SUCCESS && message) { + status = _notmuch_message_remove_filename (message, filename); + if (status == NOTMUCH_STATUS_SUCCESS) + _notmuch_message_delete (message); + else if (status == NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID) + _notmuch_message_sync (message); + + notmuch_message_destroy (message); + } + + return status; +} + +notmuch_status_t +notmuch_database_find_message_by_filename (notmuch_database_t *notmuch, + const char *filename, + notmuch_message_t **message_ret) +{ + void *local; const char *prefix = _find_prefix ("file-direntry"); char *direntry, *term; Xapian::PostingIterator i, end; - Xapian::Document document; notmuch_status_t status; - status = _notmuch_database_ensure_writable (notmuch); - if (status) - return status; - - db = static_cast (notmuch->xapian_db); + if (message_ret == NULL) + return NOTMUCH_STATUS_NULL_POINTER; - status = _notmuch_database_filename_to_direntry (local, notmuch, - filename, &direntry); - if (status) - return status; + /* return NULL on any failure */ + *message_ret = NULL; - term = talloc_asprintf (notmuch, "%s%s", prefix, direntry); + local = talloc_new (notmuch); - find_doc_ids_for_term (notmuch, term, &i, &end); - - for ( ; i != end; i++) { - Xapian::TermIterator j; + try { + status = _notmuch_database_filename_to_direntry ( + local, notmuch, filename, NOTMUCH_FIND_LOOKUP, &direntry); + if (status || !direntry) + goto DONE; - document = find_document_for_doc_id (notmuch, *i); + term = talloc_asprintf (local, "%s%s", prefix, direntry); - document.remove_term (term); + find_doc_ids_for_term (notmuch, term, &i, &end); - j = document.termlist_begin (); - j.skip_to (prefix); + if (i != end) { + notmuch_private_status_t private_status; - /* Was this the last file-direntry in the message? */ - if (j == document.termlist_end () || - strncmp ((*j).c_str (), prefix, strlen (prefix))) - { - db->delete_document (document.get_docid ()); - status = NOTMUCH_STATUS_SUCCESS; - } else { - db->replace_document (document.get_docid (), document); - status = NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID; + *message_ret = _notmuch_message_create (notmuch, notmuch, *i, + &private_status); + if (*message_ret == NULL) + status = NOTMUCH_STATUS_OUT_OF_MEMORY; } + } catch (const Xapian::Error &error) { + fprintf (stderr, "Error: A Xapian exception occurred finding message by filename: %s\n", + error.get_msg().c_str()); + notmuch->exception_reported = TRUE; + status = NOTMUCH_STATUS_XAPIAN_EXCEPTION; } + DONE: talloc_free (local); + if (status && *message_ret) { + notmuch_message_destroy (*message_ret); + *message_ret = NULL; + } return status; } -notmuch_tags_t * -_notmuch_convert_tags (void *ctx, Xapian::TermIterator &i, - Xapian::TermIterator &end) +notmuch_string_list_t * +_notmuch_database_get_terms_with_prefix (void *ctx, Xapian::TermIterator &i, + Xapian::TermIterator &end, + const char *prefix) { - const char *prefix = _find_prefix ("tag"); - notmuch_tags_t *tags; - std::string tag; + int prefix_len = strlen (prefix); + notmuch_string_list_t *list; - /* Currently this iteration is written with the assumption that - * "tag" has a single-character prefix. */ - assert (strlen (prefix) == 1); - - tags = _notmuch_tags_create (ctx); - if (unlikely (tags == NULL)) + list = _notmuch_string_list_create (ctx); + if (unlikely (list == NULL)) return NULL; - i.skip_to (prefix); - - while (i != end) { - tag = *i; - - if (tag.empty () || tag[0] != *prefix) + for (i.skip_to (prefix); i != end; i++) { + /* Terminate loop at first term without desired prefix. */ + if (strncmp ((*i).c_str (), prefix, prefix_len)) break; - _notmuch_tags_add_tag (tags, tag.c_str () + 1); - - i++; + _notmuch_string_list_append (list, (*i).c_str () + prefix_len); } - _notmuch_tags_prepare_iterator (tags); - - return tags; + return list; } notmuch_tags_t * notmuch_database_get_all_tags (notmuch_database_t *db) { Xapian::TermIterator i, end; - i = db->xapian_db->allterms_begin(); - end = db->xapian_db->allterms_end(); - return _notmuch_convert_tags(db, i, end); + notmuch_string_list_t *tags; + + try { + i = db->xapian_db->allterms_begin(); + end = db->xapian_db->allterms_end(); + tags = _notmuch_database_get_terms_with_prefix (db, i, end, + _find_prefix ("tag")); + _notmuch_string_list_sort (tags); + return _notmuch_tags_create (db, tags); + } catch (const Xapian::Error &error) { + fprintf (stderr, "A Xapian exception occurred getting tags: %s.\n", + error.get_msg().c_str()); + db->exception_reported = TRUE; + return NULL; + } }