Feeds Comment Processor provides support for creating and updating comments with the Feeds module. The Feeds module provides support for node, user, and taxonomy term imports, but does not include support for comments.

NOTE: This module requires Feeds and together they modify the database. Use on a development server and ensure your data is backed up.

• Comments
• Feeds

Install this module using the official Backdrop CMS instructions at https://backdropcms.org/guide/modules.

After enabling this module:

1. Add importer. Go to Feeds importers at admin/structure/feeds and click "Add importer". For example: Name = Comment import; Description = Comments import from CSV file.
2. Under Processor, select Comment processor.
3. Review Comment processor settings and configure as required.
4. Go to Comment processor mapping and configure source and targets, see below.
5. Configure other settings as required.

After running an import it may be helpful to check the watchdog log for messages.

The Comment ID (CID) in the source is generally not mapped to the CID in the target except when importing to existing comments. When importing new comments let Backdrop assign the CID with automatic incrementing.

However, the source CID can be mapped to GUID, and if the source includes the Parent CID for threaded comments this can be mapped to Parent ID by GUID. As long as child comments are in a later row and the Parent CID is empty when not used Backdrop automatically calculates the correct thread position.

CSV column -> Maps to target
CID -> GUID
Parent CID -> Parent ID by GUID
NID -> Node ID
Comment Title -> Subject
Body -> Comment body
etc. -> etc.

When new comments are imported or comments are updated with import, the search indexing indicator is updated for the corresponding nodes to ensure they are included on the next run of the search index update. This should avoid the need to completely rebuild the search index.

The Thread structure (thread) target is available for site builders who need to import pre-calculated thread position strings from another system.

It is not needed for normal CSV imports, as described above.

In threaded comment systems, each comment needs a position code (e.g. 01/, 01.01/, 01.02/) that determines where it appears in the conversation hierarchy. This is stored in the comment.thread database column.

• ✅ Use this target when migrating comments from another Drupal/Backdrop site and you need to preserve the exact comment ordering.
• ❌ Do NOT use for normal CSV imports. Backdrop automatically calculates the correct thread position from the pid (parent comment ID) you import.

Importing an incorrect or conflicting thread value can break the comment display order for an entire node. Only use this target if you:

1. Know exactly what the thread values represent
2. Are migrating from a system that uses the same threading algorithm

If you don't understand what this is, leave it unmapped.

This module does not depend on the Comment Notify module. But, if the Comment Notify module is enabled this module provides integration for it and supports mapping and importing of the comment notify subscription type and creating comment notify records.

The comment_notify integration takes the approach of mapping the notify options and using those to set the notified value assuming that where notifications are enabled they were already notified by the source system and the notified value is set to 1. This avoids the destination system resending email.

When importing comments with the comment_notify module enabled notices may appear several times in the watchdog log:

Notice: Array to string conversion in token_replace() (line 101 of /core/includes/token.inc).

Followed by an entry that notification email was sent although sending of email is suppressed by this module and does not occur. This may be triggered by specific comments and the number of entries relate to the number of previous comments in that thread subscribed to notifications.

These entries are harmless as the comments are imported correctly and the comment_notify table is correct.

The notices regarding the array to string conversion is related to the comment_notify module, see issue backdrop-contrib/comment_notify#8

In the Feeds module FeedsProcessor::process() calls entityValidate() before target mappings have been applied to the entity via setTargetElement(). This module works around the unexpected ordering by duplicating the logic in entitySave(). See backdrop-contrib/feeds#178

Backdrop CommentStorageController::preSave() unconditionally overwrites hostname with current request IP, see issue backdrop/backdrop-issues#7151.

This module includes a db_update() workaround. When a fix is committed to Backdrop core the workaround in this module will become a harmless redundancy that can be removed.

Backdrop CommentStorageController uses db_update() to maintain {node_comment_statistics}, that silently fails if the row doesn't exist (see core issue backdrop/backdrop-issues#5870). This causes comment bodies to be excluded from search indexing since comment_node_update_index() relies on this table. This module uses db_merge() to ensure the row always exists and is correct after import.

Additional documentation is located in the Wiki: https://github.com/backdrop-contrib/feeds_comment_processor/wiki/Documentation.

Bugs and Feature requests should be reported in the Issue Queue: https://github.com/backdrop-contrib/feeds_comment_processor/issues.

• izmeez.
• Seeking additional maintainers.

Ported to Backdrop CMS by izmeez with AI assistance.

Originally created for Drupal by Andrew Levine. Further adapted for Drupal by Dane Powell.

Drupal version maintained by Andrew Levine and twistor.

This project is GPL v2 software. See the LICENSE.txt file in this directory for complete text.