From: Jani Nikula <jani@nikula.org>
To: notmuch@notmuchmail.org
Subject: [PATCH 5/6] lib: parse messages only once
Date: Wed, 16 Oct 2013 22:00:12 +0300
Message-Id:
 <f499e217d35c69b3d02b9d455a6adbccf241d4f8.1381948853.git.jani@nikula.org>
In-Reply-To: <cover.1381948853.git.jani@nikula.org>
References: <cover.1381948853.git.jani@nikula.org>
In-Reply-To: <cover.1381948853.git.jani@nikula.org>
References: <cover.1381948853.git.jani@nikula.org>
Precedence: list

Make the necessary changes to only do one gmime parse pass during
indexing.
---
 lib/database.cc       |  2 +-
 lib/index.cc          | 70 +++++----------------------------------------------
 lib/message-file.c    |  9 +++++++
 lib/notmuch-private.h | 15 +++++++++--
 4 files changed, 29 insertions(+), 67 deletions(-)

diff --git a/lib/database.cc b/lib/database.cc
index 45a3987..d097dda 100644
--- a/lib/database.cc
+++ b/lib/database.cc
@@ -1996,7 +1996,7 @@ notmuch_database_add_message (notmuch_database_t *notmuch,
 	    date = notmuch_message_file_get_header (message_file, "date");
 	    _notmuch_message_set_header_values (message, date, from, subject);
 
-	    ret = _notmuch_message_index_file (message, filename);
+	    ret = _notmuch_message_index_file (message, message_file);
 	    if (ret)
 		goto DONE;
 	} else {
diff --git a/lib/index.cc b/lib/index.cc
index 78c18cf..71397da 100644
--- a/lib/index.cc
+++ b/lib/index.cc
@@ -425,63 +425,15 @@ _index_mime_part (notmuch_message_t *message,
 
 notmuch_status_t
 _notmuch_message_index_file (notmuch_message_t *message,
-			     const char *filename)
+			     notmuch_message_file_t *message_file)
 {
-    GMimeStream *stream = NULL;
-    GMimeParser *parser = NULL;
-    GMimeMessage *mime_message = NULL;
+    GMimeMessage *mime_message;
     InternetAddressList *addresses;
-    FILE *file = NULL;
     const char *from, *subject;
-    notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
-    static int initialized = 0;
-    char from_buf[5];
-    bool is_mbox = false;
-    static bool mbox_warning = false;
-
-    if (! initialized) {
-	g_mime_init (GMIME_ENABLE_RFC2047_WORKAROUNDS);
-	initialized = 1;
-    }
-
-    file = fopen (filename, "r");
-    if (! file) {
-	fprintf (stderr, "Error opening %s: %s\n", filename, strerror (errno));
-	ret = NOTMUCH_STATUS_FILE_ERROR;
-	goto DONE;
-    }
-
-    /* Is this mbox? */
-    if (fread (from_buf, sizeof (from_buf), 1, file) == 1 &&
-	strncmp (from_buf, "From ", 5) == 0)
-	is_mbox = true;
-    rewind (file);
 
-    /* Evil GMime steals my FILE* here so I won't fclose it. */
-    stream = g_mime_stream_file_new (file);
-
-    parser = g_mime_parser_new_with_stream (stream);
-    g_mime_parser_set_scan_from (parser, is_mbox);
-
-    mime_message = g_mime_parser_construct_message (parser);
-
-    if (is_mbox) {
-	if (!g_mime_parser_eos (parser)) {
-	    /* This is a multi-message mbox. */
-	    ret = NOTMUCH_STATUS_FILE_NOT_EMAIL;
-	    goto DONE;
-	}
-	/* For historical reasons, we support single-message mboxes,
-	 * but this behavior is likely to change in the future, so
-	 * warn. */
-	if (!mbox_warning) {
-	    mbox_warning = true;
-	    fprintf (stderr, "\
-Warning: %s is an mbox containing a single message,\n\
-likely caused by misconfigured mail delivery.  Support for single-message\n\
-mboxes is deprecated and may be removed in the future.\n", filename);
-	}
-    }
+    mime_message = notmuch_message_file_get_mime_message (message_file);
+    if (! mime_message)
+	return NOTMUCH_STATUS_FILE_NOT_EMAIL; /* more like internal error */
 
     from = g_mime_message_get_sender (mime_message);
 
@@ -502,15 +454,5 @@ mboxes is deprecated and may be removed in the future.\n", filename);
 
     _index_mime_part (message, g_mime_message_get_mime_part (mime_message));
 
-  DONE:
-    if (mime_message)
-	g_object_unref (mime_message);
-
-    if (parser)
-	g_object_unref (parser);
-
-    if (stream)
-	g_object_unref (stream);
-
-    return ret;
+    return NOTMUCH_STATUS_SUCCESS;
 }
diff --git a/lib/message-file.c b/lib/message-file.c
index 9d5a3b9..7ab9e9d 100644
--- a/lib/message-file.c
+++ b/lib/message-file.c
@@ -247,6 +247,15 @@ mboxes is deprecated and may be removed in the future.\n", message->filename);
     return NOTMUCH_STATUS_SUCCESS;
 }
 
+GMimeMessage *
+notmuch_message_file_get_mime_message (notmuch_message_file_t *message)
+{
+    if (! message->parsed)
+	return NULL;
+
+    return message->message;
+}
+
 /* return NULL on errors, empty string for non-existing headers */
 const char *
 notmuch_message_file_get_header (notmuch_message_file_t *message,
diff --git a/lib/notmuch-private.h b/lib/notmuch-private.h
index 7277df1..048dd6c 100644
--- a/lib/notmuch-private.h
+++ b/lib/notmuch-private.h
@@ -46,6 +46,8 @@ NOTMUCH_BEGIN_DECLS
 
 #include <talloc.h>
 
+#include <gmime/gmime.h>
+
 #include "xutil.h"
 #include "error_util.h"
 
@@ -320,9 +322,11 @@ notmuch_message_get_author (notmuch_message_t *message);
 
 /* index.cc */
 
+typedef struct _notmuch_message_file notmuch_message_file_t;
+
 notmuch_status_t
 _notmuch_message_index_file (notmuch_message_t *message,
-			     const char *filename);
+			     notmuch_message_file_t *message_file);
 
 /* message-file.c */
 
@@ -330,7 +334,6 @@ _notmuch_message_index_file (notmuch_message_t *message,
  * into the public interface in notmuch.h
  */
 
-typedef struct _notmuch_message_file notmuch_message_file_t;
 
 /* Open a file containing a single email message.
  *
@@ -377,6 +380,14 @@ void
 notmuch_message_file_restrict_headersv (notmuch_message_file_t *message,
 					va_list va_headers);
 
+/*
+ * get mime message. this is an ugly interface; maybe join index.cc
+ * and message-file.c, or move the top level indexing call to
+ * message-file.c with helpers in index.cc
+ */
+GMimeMessage *
+notmuch_message_file_get_mime_message (notmuch_message_file_t *message);
+
 /* Get the value of the specified header from the message as a UTF-8 string.
  *
  * The header name is case insensitive.
-- 
1.8.4.rc3