Matthew Wire [Sat, 29 Feb 2020 11:09:33 +0000 (11:09 +0000)]
Fix mistake in comment
colemanw [Mon, 24 Feb 2020 11:30:20 +0000 (06:30 -0500)]
Merge pull request #16621 from totten/master-setup-autoload
dev/core#1615 - Fix for recently migrated civicrm-setup
Eileen McNaughton [Mon, 24 Feb 2020 05:27:08 +0000 (18:27 +1300)]
Merge pull request #16622 from seamuslee001/test_dev_core_1386
dev/core#1386 Add in unit test for #15834
Seamus Lee [Mon, 24 Feb 2020 03:35:52 +0000 (14:35 +1100)]
dev/core#1386 Add in unit test for #15834
Tim Otten [Mon, 24 Feb 2020 00:37:10 +0000 (16:37 -0800)]
dev/core#1615 - Fix for recently migrated civicrm-setup
Overview
--------
Fix a recent (intra-alpha) regression from https://github.com/civicrm/civicrm-core/pull/16618
The search for `civicrm-setup` goes slightly awry because the migrated copy of civicrm-setup uses
the core's classloader rather than its own classloader.
Steps to Reproduce
------------------
* Create a site with WP and Civi(master).
* Ensure Civi is uninstalled. (ex: `cv core:uninstall -f`)
* In `wp-content/plugins/civicrm/civicrm`, touch the file `.use-civicrm-setup`.
(*This is supposed to activate the new UI.*)
* Go to the WP admin console. In the left navbar, choose "CiviCRM Installer".
Before
------
It fails to load the new UI -- the search finds `.use-civicrm-setup` but does not
find `civicrm-setup-autoload.php`. It falls back to the old UI.
After
-----
It opens the new UI.
Eileen McNaughton [Sun, 23 Feb 2020 23:25:36 +0000 (12:25 +1300)]
Merge pull request #16618 from totten/master-setup
dev/core#1615 - Import civicrm-setup
colemanw [Sun, 23 Feb 2020 20:44:47 +0000 (15:44 -0500)]
Merge pull request #16619 from demeritcowboy/remove-noop-translations
dev/translation#37 - Remove hardcoded translations from 13 years ago that no longer do anything
demeritcowboy [Sun, 23 Feb 2020 03:24:44 +0000 (22:24 -0500)]
remove hardcoded translations that don't actually do anything anyway
Tim Otten [Sun, 23 Feb 2020 00:39:21 +0000 (16:39 -0800)]
(NFC) Update civicrm-setup to current civilint
Tim Otten [Sat, 22 Feb 2020 23:55:46 +0000 (15:55 -0800)]
(dev/core#1615) Load local copy of 'civicrm-setup' instead of downloading package
Tim Otten [Sat, 22 Feb 2020 23:52:17 +0000 (15:52 -0800)]
(dev/core#1615) Import civicrm-setup code
Note: This does not import the docs. Those should go to the dev guide.
colemanw [Sat, 22 Feb 2020 23:11:14 +0000 (18:11 -0500)]
Merge pull request #16616 from totten/master-scn
Move sequentialcreditnotes under `ext/` folder
Tim Otten [Sat, 22 Feb 2020 04:20:26 +0000 (20:20 -0800)]
Move sequentialcreditnotes under `ext/` folder
Maybe this is a bit of bikeshedding, but..
* `civicrm/ext/` is used for similar content in a few other cases (eg when adding `iatspayment` to the tarball; eg for
the `drupal-case` and `wp-case` build-types).
* `civicrm/ext/` looks analgous to `sites/default/files/civicrm/ext`, which has similar content.
* Among the list of folders under `civicrm/`, `ext/` looks more consistent. There are only two
folders with capital letters (`CRM/` `Civi/`) and only two folders with multiple
words (`bower_components/` `release-notes/`)... and 18 other folders which are
lowercase+single-word (`ang/` `api/` `bin/` `css/` `distmaker/`
`extern/` `i/` `install/` `js/` `l10n/` `packages/` `settings/` `sql/`
`templates/` `tests/` `tools/` `vendor/` `xml/`).
* The file-hierarchy is additive (combining elements of the OS/host FHS, of the CMS, and of intra-extension content)
and makes for a long total.
Seamus Lee [Fri, 21 Feb 2020 21:32:00 +0000 (08:32 +1100)]
Merge pull request #16613 from seamuslee001/nfc_custom_field_array
[NFC] Convert Custom Field BAO file to use short array syntax
Seamus Lee [Fri, 21 Feb 2020 20:09:42 +0000 (07:09 +1100)]
[NFC] Convert Custom Field BAO file to use short array syntax
Seamus Lee [Fri, 21 Feb 2020 19:37:31 +0000 (06:37 +1100)]
Merge pull request #15834 from francescbassas/patch-16
dev/core#1386 Allow advanced search for contributions without a soft credit related
Seamus Lee [Fri, 21 Feb 2020 19:27:34 +0000 (06:27 +1100)]
Merge pull request #16608 from civicrm/5.23
5.23
Seamus Lee [Fri, 21 Feb 2020 10:16:26 +0000 (21:16 +1100)]
Merge pull request #16607 from totten/5.23-pathurltest
Update recently-added PathUrlTest to be more representative
Tim Otten [Fri, 21 Feb 2020 08:34:40 +0000 (00:34 -0800)]
Update recently-added PathUrlTest to be more representative
Before
------
The test generates the URL for an example page (`civicrm/mailing/subscribe`)
by calling `url()`. It does not explicitly set the `url(...$frontend...)` option,
so it defaults to `FALSE`.
By contrast, the other places which make URLs for `civicrm/mailing/subscribe`
(eg `CRM/Mailing/BAO/Mailing.php` and `CRM/Utils/Token.php`) do set `$frontend` to `TRUE`.
After
-----
The test behaves the same way as other code.
Seamus Lee [Fri, 21 Feb 2020 06:53:00 +0000 (17:53 +1100)]
Merge pull request #16545 from eileenmcnaughton/part_pend
[REF] Extract function to getTransactionInfo
Seamus Lee [Fri, 21 Feb 2020 06:49:14 +0000 (17:49 +1100)]
Merge pull request #16606 from ejegg/API4Regex-5.23
Fix two more php-finding regexes
Elliott Eggleston [Fri, 21 Feb 2020 01:34:13 +0000 (20:34 -0500)]
Fix two more php-finding regexes
Before: Civi code living under a path containing the characters php
(such as ~/src/php/civicrm) will fail to load the API4 explorer with
a message like the following:
ReflectionException: "Class \Civi\Api4\src does not exist"
After: API4 explorer loads no matter what the enclosing path
Very similar to PR #16377
Matthew Wire [Thu, 20 Feb 2020 23:06:02 +0000 (23:06 +0000)]
Merge pull request #16605 from seamuslee001/master
[REF] Update civicrm_generated following merge of #16362
Seamus Lee [Thu, 20 Feb 2020 21:15:56 +0000 (08:15 +1100)]
Merge pull request #16604 from civicrm/5.23
5.23
Seamus Lee [Thu, 20 Feb 2020 21:07:36 +0000 (08:07 +1100)]
[REF] Update civicrm_generated following merge of #16362
Seamus Lee [Thu, 20 Feb 2020 20:34:47 +0000 (07:34 +1100)]
Merge pull request #16602 from mattwire/apiv3paymentspec
Add description to params for api3 Payment.get
Matthew Wire [Thu, 20 Feb 2020 17:12:50 +0000 (17:12 +0000)]
Add description to params for api3 Payment.get
Yashodha Chaku [Thu, 20 Feb 2020 16:17:54 +0000 (21:47 +0530)]
Merge pull request #16600 from mattwire/casesettingstpl
Remove hardcoded settings from form and use SettingForm.tpl for Case settings
Matthew Wire [Thu, 20 Feb 2020 14:50:13 +0000 (14:50 +0000)]
Merge pull request #16362 from eileenmcnaughton/pp_disable
Do not enable core payment processor types that we believe likely don't work on new installs
Matthew Wire [Thu, 20 Feb 2020 13:36:52 +0000 (13:36 +0000)]
Remove hardcoded settings from form and use SettingForm.tpl for Case settings
Mathieu Lu [Thu, 20 Feb 2020 13:27:24 +0000 (08:27 -0500)]
Merge pull request #16599 from seamuslee001/dev_translation_36
dev/translation#36 Ensure that we are only passing integers to the ng…
Matthew Wire [Thu, 20 Feb 2020 12:39:19 +0000 (12:39 +0000)]
Merge pull request #16562 from eileenmcnaughton/notice
Move determination of year & month to formatCreditCardDetails
Seamus Lee [Thu, 20 Feb 2020 01:53:07 +0000 (12:53 +1100)]
dev/translation#36 Ensure that we are only passing integers to the ngettext function for plurals
Eileen McNaughton [Wed, 19 Feb 2020 22:52:54 +0000 (11:52 +1300)]
Merge pull request #16596 from civicrm/5.23
5.23
Seamus Lee [Wed, 19 Feb 2020 21:48:24 +0000 (08:48 +1100)]
Merge pull request #16594 from alifrumin/rn-5.23
First pass at release notes for 5.23
Alice Frumin [Wed, 19 Feb 2020 20:43:16 +0000 (15:43 -0500)]
First pass at release notes for 5.23
Eileen McNaughton [Wed, 19 Feb 2020 20:06:07 +0000 (09:06 +1300)]
Merge pull request #16529 from mattwire/altermailing_templatetype
Pass template_type through to alterMailing hook
francescbassas [Wed, 19 Feb 2020 17:05:35 +0000 (18:05 +0100)]
dev/core#1386 Allow advanced search for contributions without a soft credit related
Matthew Wire [Wed, 19 Feb 2020 16:25:51 +0000 (16:25 +0000)]
Merge pull request #16085 from seamuslee001/dev_core_1466
[NFC] dev/core#1466 Update Documentation URLS to be the correct links…
Matthew Wire [Wed, 19 Feb 2020 16:18:55 +0000 (16:18 +0000)]
Merge pull request #15940 from seamuslee001/dev_financial_105
dev/financial#105 Add CSS class onto the radio button payment processor options
Matthew Wire [Wed, 19 Feb 2020 16:16:56 +0000 (16:16 +0000)]
Merge pull request #15316 from mepps/order-campaign-dashboard
Secondarily order campaign dashboard by id
Matthew Wire [Wed, 19 Feb 2020 16:14:17 +0000 (16:14 +0000)]
Merge pull request #15884 from kainuk/issue-lab-1365
Related / Inherited Memberships: Custom fields not filled with data
Matthew Wire [Wed, 19 Feb 2020 16:10:04 +0000 (16:10 +0000)]
Merge pull request #16359 from jensschuppe/dev_core_1547
dev/core#1547 Add DT_RowClass CSS classes to nested group markup
Seamus Lee [Wed, 19 Feb 2020 10:08:25 +0000 (21:08 +1100)]
Merge pull request #16005 from magnolia61/Contribution_Invoice_Privacy
dev/core#1436 Do not CC or BCC Contribution invoice
Eileen McNaughton [Tue, 18 Feb 2020 21:49:59 +0000 (10:49 +1300)]
Merge pull request #16580 from civicrm/5.23
5.23
Seamus Lee [Tue, 18 Feb 2020 19:57:08 +0000 (06:57 +1100)]
Merge pull request #16588 from eileenmcnaughton/523partnotice
Fix enotice regression (unreleased)
Eileen McNaughton [Tue, 18 Feb 2020 19:47:00 +0000 (08:47 +1300)]
Merge pull request #16589 from colemanw/copyValues
Cleanup copyValues DAO function
Coleman Watts [Tue, 18 Feb 2020 13:29:17 +0000 (08:29 -0500)]
Cleanup copyValues DAO function
The $serializeArrays parameter was transitional and is now always enabled.
Passing $params by reference was unnecessary and has been removed.
Renamed a few variables for readability.
Yashodha Chaku [Tue, 18 Feb 2020 11:21:13 +0000 (16:51 +0530)]
Merge pull request #16585 from eileenmcnaughton/dep_it
Add deprecation notice
Yashodha Chaku [Tue, 18 Feb 2020 10:04:07 +0000 (15:34 +0530)]
Merge pull request #16581 from eileenmcnaughton/nfc
[NFC] Test cleanup.
eileen [Tue, 18 Feb 2020 07:23:13 +0000 (20:23 +1300)]
Fix enotice regressionn
This was merged to 5.23 but I made a booboo in it
https://github.com/civicrm/civicrm-core/commit/
a3aed42a453009733e8feb2c56fde5750b37cd14
I think the params are actually not used so it's just the notice it causes but this fixes
Seamus Lee [Tue, 18 Feb 2020 06:52:24 +0000 (17:52 +1100)]
Merge pull request #16587 from seamuslee001/5.23
Add release-notes/5.22.1.md
Eileen McNaughton [Tue, 18 Feb 2020 06:52:13 +0000 (19:52 +1300)]
Merge pull request #16575 from colemanw/savedSearch
Make savedSearch bao sane
Eileen McNaughton [Tue, 18 Feb 2020 06:51:57 +0000 (19:51 +1300)]
Merge pull request #16583 from totten/5.23-gencode-no-l10n
xml/GenCode.php - Fix execution in default locale (without l10n data)
Tim Otten [Tue, 18 Feb 2020 06:26:22 +0000 (22:26 -0800)]
Add release-notes/5.22.1.md
eileen [Tue, 18 Feb 2020 05:31:18 +0000 (18:31 +1300)]
Add deprecation notice
I've been trying to figure out when this line would be hit but I just can't find a way to trigger
it - removing it would help a lot with this code but let's deprecate for a bit.
eileen [Tue, 18 Feb 2020 04:26:06 +0000 (17:26 +1300)]
Further test fix
Seamus Lee [Tue, 18 Feb 2020 04:29:01 +0000 (15:29 +1100)]
Merge pull request #16582 from eileenmcnaughton/ids
(NFC) Remove `$Id$` from header
Coleman Watts [Mon, 17 Feb 2020 21:40:35 +0000 (16:40 -0500)]
Make savedSearch bao sane
Tim Otten [Tue, 18 Feb 2020 03:40:14 +0000 (19:40 -0800)]
xml/GenCode.php - Fix execution in default locale (without l10n data)
Overview
--------
This fixes a recent regression in which `xml/GenCode.php` fails to execute in certain configurations.
Initial report: https://github.com/civicrm/civicrm-buildkit/issues/503
Before
------
* If the folder `l10n` exists in its traditional location, and if you run `./bin/setup.sh -g`, then it
correctly executes.
* If the folder `l10n` does not exist in its traditional locacation, and if you run `./bin/setup.sh -g`,
then it fails with an error like this:
```
+ php -d mysql.default_host=127.0.0.1:3306 -d mysql.default_user=dmasterciv_abcde -d mysql.default_password=
abcd1234 GenCode.php schema/Schema.xml '' Drupal
civicrm_domain.version := 5.23.alpha1
Parsing schema description schema/Schema.xml
Extracting database information
Extracting table information
PHP Fatal error: Uncaught RuntimeException: Invalid configuration: the [cms.root] path must be an absolute URL in /home/me/.../Civi/Core/Paths.php:174
Stack trace:
#0 /home/me/.../Civi/Core/Paths.php(151): Civi\Core\Paths->toAbsoluteUrl('/', 'cms.root')
#1 /home/me/.../Civi/Core/Paths.php(176): Civi\Core\Paths->getVariable('cms.root', 'url')
#2 /home/me/.../Civi/Core/Paths.php(151): Civi\Core\Paths->toAbsoluteUrl('/', 'civicrm.root')
#3 /home/me/.../Civi/Core/Paths.php(224): Civi\Core\Paths->getVariable('civicrm.root', 'path')
#4 /home/me/.../Civi/Core/Paths.php(84): Civi\Core\Paths->getPath('l10n')
#5 [internal function]: Civi\Core\Paths->Civi\Core\{closure}()
#6 /home/jo in /home/me/.../Civi/Core/Paths.php on line 174
```
After
-----
You can run `./bin/setup.sh -g` with or without the `l10n` folder.
Comments
--------
There's an argument to be made that this shouldn't be necessary: when running
`GenCode`, it should only create PHP files, and none of the output should depend
on the CMS URL - because that's liable to change when you deploy the PHP code.
If someone did try to generate URL at such an early stage, it's arguably good to
generate an error. In point of fact, `GenCode` is not actually using the CMS
URL.
The oddity stems from the contract of `CRM_Utils_System_*` (specifically,
`getCiviSourceStorage()` and `getDefaultFileStorage()`) which do double-duty
providing both path and URL. To avoid duplicate work, `Civi\Core\Paths` uses
the same grain of information - it tracks the pairs of path+URL.
A more aggressive fix might be to split `getCiviSourceStorage()` and
`getDefaultFileStorage()` so that it's possible to get the path and URL
separately; then revise `Civi\Core\Paths` to take advantage of the finer-grained
contract. However, splitting those things would be a more invasive patch,
and we're currently in RC.
eileen [Tue, 18 Feb 2020 01:01:53 +0000 (14:01 +1300)]
[NFC] Test cleanup.
Cleaned up this test - re-used code etc
colemanw [Tue, 18 Feb 2020 02:57:39 +0000 (21:57 -0500)]
Merge pull request #16531 from eileenmcnaughton/cn_
dev/financial#84 Move sequential credit notes from 'deeply embeded functions to separate structure
Tim Otten [Tue, 18 Feb 2020 01:38:18 +0000 (17:38 -0800)]
Merge pull request #16528 from mattwire/removeoldflexmailer
Remove old defines for flexmailer that haven't been required since CiviCRM 5.x
eileen [Tue, 18 Feb 2020 01:18:29 +0000 (14:18 +1300)]
Remove from header
I was a bit surprised to see the old drupal svn value pop back up - did if come from this header.txt file
Seamus Lee [Tue, 18 Feb 2020 01:01:06 +0000 (12:01 +1100)]
Merge pull request #16578 from seamuslee001/dev_core_1602
dev/core#1602 Fix join date filtering in search builder by standardis…
Seamus Lee [Mon, 17 Feb 2020 23:00:46 +0000 (10:00 +1100)]
dev/core#1602 Fix join date filtering in search builder by standardising date field handling in Membership Query BAO
Seamus Lee [Mon, 17 Feb 2020 23:32:25 +0000 (10:32 +1100)]
Merge pull request #16563 from eileenmcnaughton/selfsvc
[NFC] Minor code cleanup
Eileen McNaughton [Mon, 17 Feb 2020 22:57:11 +0000 (11:57 +1300)]
Merge pull request #16577 from civicrm/5.23
5.23 to master
Eileen McNaughton [Mon, 17 Feb 2020 22:53:15 +0000 (11:53 +1300)]
Merge pull request #16558 from eileenmcnaughton/523dupe
dev/core#1597 & dev/core#1595 fix regression when deduping on custom fields
Sandor Semsey [Mon, 17 Feb 2020 22:11:37 +0000 (23:11 +0100)]
CRM-21541 CRM.moneyFormat can't handle values with zero or one decimals (#16445)
* CRM.formatMoney() handles zero or one decimals
* CRM.formatMoney() handles zero or one decimals
* CRM.formatMoney() handles zero or one decimals
* CRM.formatMoney() handles zero or one decimals
* CRM.formatMoney() handles zero or one decimals
* CRM.formatMoney() handles zero or one decimals
* CRM.formatMoney() handles zero or one decimals
* missing semicolon
* comment
* comment
* comment
* comment
Seamus Lee [Mon, 17 Feb 2020 21:43:07 +0000 (08:43 +1100)]
Merge pull request #16564 from eileenmcnaughton/test
dev/financial#102 Extend test to check status
eileen [Sun, 16 Feb 2020 21:26:03 +0000 (10:26 +1300)]
dev/core#1597 & dev/core#1595 fix regression when deduping on custom fields
Eileen McNaughton [Mon, 17 Feb 2020 21:13:27 +0000 (10:13 +1300)]
Merge pull request #16571 from wmortada/over-ride
(NFC) Fix typo - no hyphen in 'override'
Eileen McNaughton [Mon, 17 Feb 2020 20:58:10 +0000 (09:58 +1300)]
Merge pull request #16095 from seamuslee001/dev_core_1450
dev/core#1450 Try reducing the size of the temporary table so it only…
Eileen McNaughton [Mon, 17 Feb 2020 20:52:44 +0000 (09:52 +1300)]
Merge pull request #16574 from civicrm/5.23
5.23
Seamus Lee [Mon, 17 Feb 2020 20:50:10 +0000 (07:50 +1100)]
Merge pull request #16567 from eileenmcnaughton/parti_form
Fix theoretical regression
Seamus Lee [Mon, 17 Feb 2020 20:06:53 +0000 (07:06 +1100)]
Merge pull request #16572 from civicrm/5.23
5.23
William Mortada [Mon, 17 Feb 2020 18:04:18 +0000 (18:04 +0000)]
Fix typo - no hyphen in 'override'
Matthew Wire [Mon, 17 Feb 2020 16:01:31 +0000 (16:01 +0000)]
Merge pull request #16506 from mattwire/invoicedue_countrefunded
Count refunds when calculating amount due for an invoice
colemanw [Mon, 17 Feb 2020 15:20:06 +0000 (10:20 -0500)]
Merge pull request #16009 from jmcclelland/import-match-phone
dev/core#1438 Enable matching on contact phone when importing contributions
Matthew Wire [Mon, 17 Feb 2020 15:16:22 +0000 (15:16 +0000)]
Merge pull request #16035 from demeritcowboy/ang-crmcasetype-list-boolean
dev/core#1451 Incorrect dropdown action choices on case type listing screen
Matthew Wire [Mon, 17 Feb 2020 14:33:57 +0000 (14:33 +0000)]
Merge pull request #16565 from eileenmcnaughton/contview
dev/financial#118 Move source & receive date to the top on contribution view page
Yashodha Chaku [Mon, 17 Feb 2020 13:25:25 +0000 (18:55 +0530)]
Merge pull request #16568 from JGaunt/patch-4
Update 5.23.0.md
Matthew Wire [Mon, 17 Feb 2020 13:11:28 +0000 (13:11 +0000)]
Merge pull request #16491 from mfb/undefined-index
Resolve notices if first donation amount and date columns were disabled
Matthew Wire [Mon, 17 Feb 2020 13:03:11 +0000 (13:03 +0000)]
Merge pull request #16455 from MegaphoneJon/core-1568
core#1568: Show recipientListing to non-admins
Jade Gaunt [Mon, 17 Feb 2020 11:52:39 +0000 (11:52 +0000)]
Update 5.23.0.md
Added organisation name :)
Matthew Wire [Mon, 17 Feb 2020 11:01:19 +0000 (11:01 +0000)]
Merge pull request #14928 from lcdservices/dev-core-1158
dev/core#1158 mailing labels: explicitly set primary flag
Matthew Wire [Mon, 17 Feb 2020 08:46:03 +0000 (08:46 +0000)]
Merge pull request #15856 from agileware/CIVICRM-1368
dev/financial#104 Use rounding and integers to compare monetary values…
eileen [Mon, 17 Feb 2020 08:04:00 +0000 (21:04 +1300)]
Fix theoretical regression
This PR https://github.com/civicrm/civicrm-core/pull/16363/commits/
44237a295dc5eab4f6110fb0556b1af742364f6e
(merged to 5.23) calls self::::getDefaultRoleID() in a way that was always flawed but definitely woudl
not work now. I think it is never reached but better to put this fix into 5.23 & maybe reemove the line in master
later since it assumes participant_status_id is optional - which is not true of the backoffice form
eileen [Mon, 17 Feb 2020 06:28:43 +0000 (19:28 +1300)]
dev/financial#118 Move source & receive date to the top on contribution view page
Per https://lab.civicrm.org/dev/financial/issues/118 this doesn't seem to be controversial - although there is a desire
to take it further & once this is merged I'll look at that
eileen [Mon, 17 Feb 2020 06:03:03 +0000 (19:03 +1300)]
dev/financial#102 Extend test to check status
It turns out the code is working fine (fixed in the meantime?) and I just needed to extend the checks
a bit, did some tangental cleanup
eileen [Mon, 17 Feb 2020 04:50:04 +0000 (17:50 +1300)]
[NFC] Minor code cleanup
Declare exceptions, use single quotes & strict comparison, do not declare unused params
eileen [Mon, 17 Feb 2020 01:37:16 +0000 (14:37 +1300)]
Move determination of year & month to formatCreditCardDetails
Per https://github.com/civicrm/civicrm-core/pull/16514 it makes sense as a better place for these
Eileen McNaughton [Mon, 17 Feb 2020 00:34:02 +0000 (13:34 +1300)]
Merge pull request #16557 from eileenmcnaughton/find
[NFC] Preliminary cleanup
Seamus Lee [Sun, 16 Feb 2020 22:50:36 +0000 (09:50 +1100)]
Merge pull request #16498 from seamuslee001/net_smtp_upgrade
Upgrade Net_SMTP Package and remove now unneded patches and mov…
Seamus Lee [Sun, 16 Feb 2020 21:39:24 +0000 (08:39 +1100)]
Update patch URL and remove patch content as per standard now
Seamus Lee [Sat, 8 Feb 2020 03:40:42 +0000 (14:40 +1100)]
[REF] Upgrade Net_SMTP Package and remove now unneded patches and move to using composer patches rather than patching in a script file
eileen [Sun, 16 Feb 2020 21:01:51 +0000 (10:01 +1300)]
[NFC] Preliminary cleanup
Declare exceptions, params, don't declare unused variables
colemanw [Sun, 16 Feb 2020 19:36:56 +0000 (14:36 -0500)]
Merge pull request #16552 from eileenmcnaughton/part_form
[REF] Fix static call to non-static function.
eileen [Wed, 12 Feb 2020 13:23:25 +0000 (02:23 +1300)]
dev/financial#84 Move sequential credit notes from 'deeply embedeed in BAO functions' to usin a listener structure
https://lab.civicrm.org/dev/financial/issues/84
We have a significant problem in our code base that it is hard to cleanup our key processes and make them usable and
Form-builder ready because of all the features deeply interwoven into them. It will be important that our
Order->add payment flow works and can be utilsed by the form builder. The expectation is that adding a payment
will update all the appropriate entities. We've been working pretty hard on this for a couple of years now (I think the time
taken to clean up code is sadly greater than the time it takes to add it :-() and the Payment.create part is getting
pretty solid now.
However, we are consistently hampered by code that has been 'enmeshed' into deep BAO code. In the case of the credit note
creation something that can easily be implemented as a pre hook and kept entirely out of the Contribution.create function
is not only embedded in there but also in updateFinancialAccountsOnContributionStatusChange - which is one of the functions
we are some dozens of hours into adding tests & 'sorting out'. We need to find a way to
1) Declutter key functions where possible and
2) Model good ways of implementing things by hook (& work through limitations in this approach) as part of cleaning
up core and pushing towards a LeXIM approach.
I originally agreed with the CiviDiet approach that we should try to move the myriad of things that are 'add ons' that someone
thought was necessary & useful to others (or less charitably they thought that if they got it into core someone else would have
to maintain it for them) moved out of core into extensions and eventually the maintenance of them moved out of the core team too.
However, it's clear from the discussion on moving credit notes out of core that this is too hard for the forseeable future. We can try to avoid more features being added to core but most features in core are used by at least one organization & the politics of
moving things out are too hard. On the bright side we let a lot less new features into core these days - not none - the membership status override date springs to mind and the ongoing adding of fields & filters to reports - but less.
This leaves us with the question of whether we can still achieve the goal of making add-on features less of a developer burden.
In addition to ensuring code is tested we can do this by getting it out of the 'deep functions' into a listener structure.
I was able to covert the 'spaghetti-like-tendrils' in the credit note code to a simple 'pre' hook.
So it felt like we needed to create something new. Something like .. an extension .. but not visible on the extensions screen
and in core. Something like ... an extension in a folder in the core codebase that is suppressed from the extensions UI....
So this adds it as an extension to core in a new location & ensures GUI users cannot see it. I don't know exactly where we
draw the line around packaging things this way - Partial payments is clearly part of the core architecture, but there might be
a world in which we move something like event cart to be a more distinct bundle. As we move to form builder forms we are going to
want forms being replaced to be able to be 'quarantined' for future amputation.... I would be looking to move financial type acls and
invoicing to this structure - mostly because they are the biggest 'offenders' in terms of having code all over the place causing
problems with other code initiatives & we've struggled to come up with a way to deal with that other than putting touching that
code in the 'too hard basket'. But even things like PCPs & Premiums would be more maintainable if they were discrete & overcoming
the challenges involved in that would solve a lot of extensibility issues.
Note this approach permits an incremental cleanup-up-to-separation process. For example we could create an extension for a
feature and move some code to it and later move some other code. For example we could move the invoicingPostProcessHook
to a shell invoicing extension and at some later point address other parts of the invoicing code (check LineItem::getLineItems
for the chunk of code that has been blocking other cleanup - moving this to a listener would help a lot
Note this PR does not address the significant performance issues associated with the creditnotes code on larger sites, only the impact
it has on our code maintainability. If we can agree a way to start to address how to split out or code into listeners in an agreed structure then I can look at that as a follow up.