| Age | Commit message (Collapse) | Author |
|---|
|
Since the image util method `image_base64()` is converting BMP images to PNG
automatically[1], there would be a mismatch between the attachment content type
and the attachment mimetype.
- Content type: retrieve with `base64_to_image(data).format` -> return PNG as
the image data was converted from BMP to PNG
- Mimetype: set on `write()` and `create()` with `_check_contents()` which
call `mimetypes.guess_type(file_name)`.
BMP image `my_file_image.bmp` would have a mimetype set to BMP
The computed field `image_src` would return `False` since the image.mimetype
would be BMP, making the dialog picker not showing the image but loading
`http://localhost:8069/false` instead.
[1] https://github.com/odoo/odoo/blame/097f29d0f6a5b295c81fd074665dbd261eaf12fb/odoo/tools/image.py#L124
task-2345082
Coming from #65828
Closes #68326
X-original-commit: ef5b108b781df73342bb99342a1116fad86d3b58
|
|
*: website, base
The normal flow to render a template is now to use `render_public_asset`
which bypasses the read access rights if the user matches the groups
the view declares.
For public users, we still cannot use that as they do not have access
to calling model methods at all. The route `public_render_template` is
thus still needed, but it should use the `render_public_asset` util. In
master, it will only rely on it but as a stable fix, we still need to
use the whitelist as we cannot change the related views' group.
This commit also adds deprecated comments to the read_template method
which has been removed in master.
Related to task-2412544
closes odoo/odoo#64137
Signed-off-by: Quentin Smetz (qsm) <qsm@odoo.com>
|
|
Previously, when flipping some shapes vertically on chrome, and
vertically or horizontally on firefox, there was a gap of 1px or less at
some screen widths. This is caused by the way these browsers render
backgrounds on transformed elements. It seems like firefox will round
the size of the element to render the background image on it if the
element is transformed, while chrome will always render the background
at the correct size (or render it at a larger size and clip it to the
correct size after the fact), but will round the coordinates of the
symmetry point, causing only the vertical flip to have this issue, and
only if the background image would not extend beyond the bounds of the
element (on firefox, even shapes that would extend beyond the bounds of
the element would showcase a gap, eg origins 1)
This commit fixes that by applying the transform to the SVG rather than
to the element, by use of the shapes controller.
task-2369560
closes odoo/odoo#61114
Signed-off-by: Quentin Smetz (qsm) <qsm@odoo.com>
|
|
In order to allow modifying background images, we need information on
the image, this is done by creating a placeholder img and calling the
loadImageInfo function on it. However this function did not account for
the case where an image had not src attribute, which causes an
unnecessary rpc. Other problems could arise from this as an attachment
that doesn't have the correct mimetype but has a matching src could be
returned, causing its image_src field to be false, which we would then
attempt to load as a valid image, causing crashes.
This commit fixes that by not trying to load image infos when the src of
an image is empty, only looking for attachments of the supported
mimetypes, and also checking that we actually did receive an image_src
before setting it as the original src of the image, which will prevent
accidentally trying to load a falsy src as an actual image.
closes odoo/odoo#60953
Signed-off-by: Quentin Smetz (qsm) <qsm@odoo.com>
|
|
*: website, web_unsplash, website_blog
We have recently created an odoo media library that exposes a public API
for searching for open-source media, including dynamic SVGs that can be
used with the dynamic background-shape system to create illustrations
whose colors can be changed by the user. This commit adds the
corresponding client-side functionalty: search for the images in the
media-library, creating a corresponding attachment when selected, as
well as an option to customize the colors of the dynamic SVGs.
In the process, adaptations have been made in the web_unsplash module to
avoid duplicating code and to adapt code that was making assumptions
that are no longer valid.
To make new users more aware of this feature, all default images have
been hidden from the media-dialog, so that the user is prompted so
search for images. Some tests and tours that were relying on having
default images in the media-dialog (eg in website_blog) have been
adapted to match.
task-2316651
Linked to: odoo/iap-apps#235
closes odoo/odoo#56479
Signed-off-by: Jérémy Kersten (jke) <jke@openerp.com>
|
|
Issue
- Select "My Company (Chicago)" as current company
- Open "Projects" module
- Create a new Task
- Edit description, upload a new Image and add it to description.
Multi-company error "Write" on project.task .
In frontend, can optimize uploaded picture.
Causes
Wrong "allowed_company_ids" value in context.
Solution
Remove allowed companies from context to avoid allowed_company_ids
which may erroneously restrict based on website.
opw-2304511
closes odoo/odoo#55897
X-original-commit: 8f6506a57853040d065569fdd60f9d94be8e26e1
Signed-off-by: Nicolas Lempereur (nle) <nle@odoo.com>
Signed-off-by: bon-odoo <nboulif@users.noreply.github.com>
|
|
This commit adds a new snippet option allowing the user to choose one of
a series of SVG shapes as their background, or to be overlaid over the
background image of a snippet and to customize the colors of those
shapes.
Part of https://github.com/odoo/odoo/pull/53017
task-2210790
closes odoo/odoo#53017
Signed-off-by: Quentin Smetz (qsm) <qsm@odoo.com>
|
|
When clicking on an image in edit mode, the web editor attempts to find
an attachment that corresponds to the image's URL. In the case of the
website logo, the URL looks like it's referencing an attachment by id or
xml-id, but it's actually neither. This cause a variable to never be
initialized, which in turns causes a traceback when trying to access it
later.
This commit fixes that by always initializing attachment to None.
closes odoo/odoo#53418
X-original-commit: 790e33de190392b937e21c73905db4d48cfb86aa
Signed-off-by: Quentin Smetz (qsm) <qsm@odoo.com>
|
|
*: website, mass_mailing
In continuity with the nline crop, in an effort to make the image
workflow as seamless as possible in the website editor, the image
optimization feature has been moved out of the modal which could be
opened through the media-dialog and into the left panel.
This commit adds a quality slider and a width selector, as well as a
preview of the image's weight to the left panel, and removes the
image_optimize dialog.
This commit also introduces the possibility to add a color filter on
images.
Part of https://github.com/odoo/odoo/pull/51517
task-2192755
closes odoo/odoo#51517
Signed-off-by: Quentin Smetz (qsm) <qsm@odoo.com>
|
|
*: base
Previously, cropping an image was done inside of a modal window, turning
it into an inline widget makes the editor feel more integrated and
allows you to better see the result of cropping directly in the page,
it's also less intrusive to the user's edition workflow.
At the same time, the way cropped images are saved was refactored to use
the new original_id field on ir_attachment, as the old system was
creating problems.
Part of https://github.com/odoo/odoo/pull/51517
task-2192755
|
|
Previously, the media-dialog had a "filter" feature, where when changing
a background image for example, all backgrounds would show first in the
media-dialog. However, because this required to load all public
attachments in the database it was causing issues and was removed. The
backend code for this feature however was unfortunately forgotten, this
commit removes that code.
Part of https://github.com/odoo/odoo/pull/51517
|
|
render, render_template, load, activity_schedule_with_view,
get_website_pages should all be private:
It should not be possible to render an aribtrary template only with
its name or id
Still need to render some qweb views from js so the method
render_template is kept public.
This explains why the website editor still need read access on
ir.ui.view as we want to allow any snippet to be rendered.
|
|
If for some reason someone wanted to develop a textarea using the
editor which is supposed to work as a public user (like we are trying
to do on Odoo.com), it was not possible. The code was "designed" to
allow it but there was one problem: the lazy loading of the editor
assets required a `render_template` call to the server... which cannot
be done from a public user.
This commit solves the issues by allowing the lazy loading of assets
to use a custom route if required. That route is then used by the editor
"root". That route performs the render_template as a superuser provided
that the view's xmlid is whitelisted.
Note: there was another unauthorized call for public user: the
colorpicker. This was solved by disabling the colorpicker template rpc
for public user, they will still get the default summernote one.
Part of https://github.com/odoo/odoo/pull/48981
closes odoo/odoo#48981
closes odoo/odoo#49398
X-original-commit: e84a0bfdc99c21406861b88c02b11c925d92f927
Signed-off-by: Quentin Smetz (qsm) <qsm@odoo.com>
|
|
Previously, when selecting an image, the user could optimize their
images by clicking a little cog icon in the media dialog. They would
also be prompted to optimize their images when uploading them. In most
cases, it makes a lot more sense to always save the original in the
database, and optimize the original whenever it's dropped in the page
automatically.
A second problem is that when using the optimization feature, optimized
images would be displayed along the originals by default, resulting in a
lot of duplicates while browsing the media dialog.
This commit changes both of these things:
- By default, when uploading an image, the original uploaded image is
used. The user can still optimize their image manually through the
media-dialog by using the cog icon. When choosing an image in the
media-dialog, if it's not already an optimized image, an optimized copy
is automatically created and used instead.
- Optimized images are now hidden by default, in the media dialog, and
can be shown by clicking a checkbox at the top of the modal when in
debug mode
task-2091417
closes odoo/odoo#45174
|
|
This reverts commit ff1c35513aecef5fbc30f7c2f3834450e719714e.
closes odoo/odoo#41480
X-original-commit: 116057b26e71db4692280463669f3e80d813ddcc
Related: odoo/enterprise#7110
Signed-off-by: Yannick Tivisse (yti) <yti@odoo.com>
|
|
FontAwesome 5 introduced new names for some icons as described on
https://fontawesome.com/how-to-use/on-the-web/setup/upgrading-from-version-4#name-changes
This commit replaces the old names to the new ones.
closes odoo/odoo#35826
Taskid: 2050241
Related: odoo/enterprise#5180
Signed-off-by: Yannick Tivisse (yti) <yti@odoo.com>
|
|
closes odoo/odoo#37058
Signed-off-by: Nans Lefebvre (len) <len@odoo.com>
|
|
|
|
In multi-level checkboxes, if there is only one checkbox per level, the
expected behavior is that all checkbox will be checked if one of them is
checked. This was not the case before this commit.
|
|
The wysiwyg abstraction layer was introduced in saas-12.2 by
commit f2969923. Its goal was to provide an external interface
for other modules that wanted to use the editor feature such
that the editor core itself could be changed without requiring
extensive changes in the other modules.
This commit is adapting the wysiwyg interface to instantiate
the 12.0 editor below it rather than the saas-12.2 one, thus
reverting the "new editor" part of f2969923 itself.
Part of PR 35677.
Co-authored-by: Nicolas Bayet <nby@odoo.com>
Co-authored-by: Antoine Guenet <age@odoo.com>
Co-authored-by: Christophe Matthieu <chm@odoo.com>
Co-authored-by: David Monjoie <dmo@odoo.com>
|
|
* = website, website_blog
The goal is to give the user an opportunity to optimize his images before using
them.
For this we introduce a preview/configuration dialog after every image upload,
where appropriate default values are filled for the quality and resolution,
based on where the image is going to be used. Since this is not going to be
perfect all the time, we still allow the user to configure them, and we display
a preview to ease this process.
Indeed it is important for SEO and for usability in general that the images are
as light as possible in size.
Technically the original image is uploaded and saved first, and then it can be
optimized. This way the upload only happens once, and the preview can be
computed from the already saved image.
task-1930726
PR: #31208
|
|
* = web_unsplash, website
Remove iframe upload and replace it with a proper RPC. The route is now only
accepting one file at a time, which makes it easier to maintain, and allows to
process the first file as soon as it is ready, instead of having to wait on
everything.
Also fix a bug where a document could not be uploaded because it would try to
parse it as an image.
Remove filter after search result from the JS and add the corresponding
conditions in the domain to avoid processing and returning unnecessary records.
If multiple attachments exist with the same URL, now display them all to avoid
confusion.
Compute all the necessary values in the python model instead of in the JS.
Allow SVG in the widget (the server already accepted them before).
Disable multiple image upload when not in multiImages mode.
task-1930726
PR: #31208
|
|
* = web_unsplash, website_blog
Rename "image" to "attachment" when appropriate in FileWidget
Improve image selection / highlight code (now based on attachment id)
Improve image grid view:
- Remove the row mechanism and replace it by a single loop over the images.
- Take advantage of row wrap for a responsive display.
- Uniformize the height (prevent tall images from moving down everything).
- Load a miniature of the images when possible, remove the loading placeholder.
This should significantly reduce the download size when opening the dialog,
at the cost of CPU usage on the server.
- Split the existing image view into one view for the loop and one view for the
image itself. This way in a following commit we can reuse the image view
when adding a new one, instead of having to render everything again.
task-1930726
PR: #31208
|
|
* = hr, im_livechat, mail, payment, purchase, web, web_editor, web_unsplash,
website_profile, website_slides
Since the merge of all image tools into one function, the intermediary functions
are not needed anymore.
task-1958000
PR: #31811
|
|
This commit introduces the possibility to resize and to select the quality when
optimizing an image for web.
This will be used by the media dialog in a future commit.
The arbitrary resolution limit of 42e6 has been increased to 45e6 to fit some
of the biggest images we might get from Unsplash, and this value has been moved
into a variable so it can be customized.
The mimetype computation has been moved after the image processing because in a
next commit the operation could return the image in a different format than what
was originally given. Eg. BMP -> PNG, or unsupported format -> JPEG.
task-1958000
PR: #31811
|
|
Commit https://github.com/odoo/odoo/commit/29075a8e2c734624cec4f6009185616634f70dea
introduced some python methods which became useless with commit
https://github.com/odoo/odoo/commit/aef6a5aaed22b250799ce92b5021d240fed9e12e
They were not removed with that commit since it was a stable version and
because it would have made forward-port difficult because of master
tasks. Now that those tasks are merged, this commit removes the useless
methods.
|
|
* web_editor
This commit's goal is to optimize the delay to make a scss customization
thanks to the website customize dialog. Before this commit, the JS code
took advantage of existing web_editor routes but this was far from
efficient as this was done:
1) PY: Load all scss files used in the page
2) JS: Find the one we want to customize and adapt the scss content
3) PY: Create the scss customization with the new content
(So two RPC (including one very slow) and content building in
javascript)
After this commit, this is done:
1) PY: Load the scss file being customized, adapt its content according
to the new values and save the customization
(So the whole logic is in python, requiring only one small RPC)
This commit also take the opportunity to review the web_editor ace
editor loading code (the logic is shared with website customizations).
PR https://github.com/odoo/odoo/pull/29624
task-1904244
|
|
* web_editor
Currently, when an user wants to edit the css of its website, he has
access to lots of files and, in debug mode, to even more files. While
the feature is nice for people who know what they are doing, a lambda
person can easily break its website by changing the wrong files. Also,
nothing warns the user that the files that are edited will never receive
updates anymore (unless if they are reset). The goal is to prevent
those behavior while still allowing the add custom css. Another goal
is to be able to *add* custom javascript.
Custom files created for the user:
SCSS: user_custom_rules.scss, user_custom_bootstrap_overridden.scss
JS: user_custom_javascript.js
See https://github.com/odoo/odoo/pull/29999
task-1919350
|
|
|
|
|
|
Before this commit:
1. If we were editing a template with no xml_id, typically the case when
editing a specific view (COW'd), the 'Template ID' label would be left empty
as the view does not have an xml_id but only a key.
2. If we were editing a generic view, the template would be COW'd and replaced
by the specific one created during the save.
After the save, the page would be reload and will try to reopen the edited
template (stored in url `res=123`). But as this template would be replaced
by the COW'd view, the JS would crash trying to access an inexisting view.
The HTML editor would then open in a very thin modal, barely editable.
Now:
1. Display the template's key instead of its xml_id. The key is basically a
duplicate of the xml_id, mainly used in website.
2. If we are saving a generic view, we search for its newly created specific
view to upload the URL hash before reloading.
task-1934279
Fix #30447
|
|
|
|
|
|
|
|
|
|
closes odoo/odoo#30566
|
|
Modifying less files and saving them using the web editor can raise
and error. The error is due to the way odoo save the modified file.
It extends the view containing the URL of the file by filtering views
related to the xml bundle id and keeping those that have the said URL
in their content.
This lookup fails to give just one result in case there are multiple
file containing the said URL. This fix change the filter from a
bare `if url in view.arch` to a correct xpath lookup.
The "/web/static/src/variables.less" less file was one of the
problematic files.
opw-1889794
closes odoo/odoo#30186
|
|
Broken by https://github.com/odoo/odoo/pull/30186
closes odoo/odoo#30403
|
|
|
|
|
|
While creating a new record, the default value of a field used with the
'html_frame' field widget was not used.
Discovered with task-1903256
closes odoo/odoo#28838
|
|
Odoo no longer supports python 2, thus some of these helpers can and
have been replaced by python 3 built-ins, therefore there is no need for
them to stay defined.
The removed helpers are:
* izip, imap and ifilter
* unichr, text_type
* implements_to_string, implements_iterator
* string_types, integer_types
* to_native
The python 2 shims have also been removed, and only the python 3 helpers
have been kept, because they can still be usable (i.e. accepting
both bytes and str for functions that can only accept one of the two)
[REM] pyjsparser: remove PY3 shims
They're no longer necessary as Odoo doesn't officially support python 2
anymore.
closes odoo/odoo#28519
|
|
This commit replaces calls to pycompat helpers that were intended for
python 2 <-> python 3 interoperability for python 3 builtins, as python
2 is no longer officially supported by Odoo.
This includes:
* calls to imap/izip/ifilter replaced by map/zip/filter
* uses of text_type replaced by str
* uses of unichr replaced by chr
* calls to implements_to_string, implements_iterator removed
* string_types and integer_types replaced by str, int respectively
* calls to to_native replaced by calls to to_text
This is done in preparation to the removal of these deprecated helpers
in the following commit.
|
|
closes odoo/odoo#30614
|
|
* Creating a new structure by transforming all the plugins in the
library using the odoo inheritance system. Plugins are easier to
implement with the AbstractPlugin to add Odoo behaviors.
* From now on, the methods of the library (in this case Summernote) can
no longer be called by other modules or files. Only the wysiwyg
widgets can access it, to simplify the updating process. The wysiwyg
object serves as an interface.
* Depending on the options the snippets will be loaded or not, the
editor will be in an iframe or not... all of this is transparent from
the outside.
* Regarding iframes, all controllers related to editing have been
removed: the new API no longer needs them. This speeds up loading,
eases testing and removes complexity for the same
features.
PUBLIC FEATURES
There are several public methods on the Wysiwyg class:
* Wysiwyg.prepare (WidgetParent): returns a deferred resolved when the
library (xml, lazy, assets...) is loaded.
* Wysiwyg.getRange (DOM): returns the range (selection in the dom)
* Wysiwyg.setRange (startNode, startOffset, endNode, endOffset): creates
a range (selection in the dom)
* Wysiwyg.setRangeFromNode (DOM, options) that creates a range from an
element (option available to select all, start or end)
A jQuery selector was added: :o_editable, which indicates whether the
current element is editable. That is, if it is contained in a tag with
the attribute 'contentEditable = "true"' or in a tag with the class
o_editable.
Several methods are also present:
* focusIn: makes a focus and places the cursor at the beginning of the
element
* focusInEnd: makes a focus and places the cursor at the end of the
element
* selectContent: makes a focus and selects the content
HTML FIELD
The HTML field can receive different options:
* style-inline: {boolean} transforms a class into an inline style when
saving and vice versa when reading.
* no-attachment: {boolean} prevents the use of attachments (in media
dialog)
* cssEdit: {xml_id} to use a template containing the css to loaded in
an iframe when editing
* cssReadonly: {xml_id} to use a template containing the css to load
into an iframe when viewing in readonly
* snippets: {xml_id} snippets template (can be used with or without
cssEdit)
* wrapper: {template} qweb static template (containing a tag:
id = "wrapper") that will include the content during editing (removed
on save)
MASS MAILING
A widget was created for mass mailing. There are now two fields:
body_html and body_arch.
body_arch contains the code with the class without conversion into
inline style, useful when editing and one with the inline style that is
visible in readonly mode and sent by email.
Advantage: no spreading errors, able to update css/theme, able to do
more changes when converting to inline style so that a maximum of mail
clients have an impeccable rendering.
Co-authored-by: Antoine Guenet <age@odoo.com>
|
|
* base, website
An image URL may be like "/web/image/<xml_id>" where <xml_id> is the
xml id of an ir.attachment. In fact, the route does not use the xml_id
but the 'key' field, which is normally equal to the xml_id for standard
attachment. In the context of themes + multi-website, the value used as
an xml_id may refer to a theme.ir.attachment instead of an ir.attachment
and this led to a bug as those must not be used the same way.
The crop route which gets image information is adapted by this commit to
always look for the *current website* *ir.attachment* related to the
given xml_id/key.
Thanks to @kig-odoo for spotting the issue.
Closes https://github.com/odoo/odoo/pull/27588
closes odoo/odoo#27611
|
|
* web_editor, website_blog, website_crm_partner_assign, website_links,
website_livechat, website_mail_channel, website_mass_mailing,
website_slides
Calling a model's method thanks to an RPC will now always send the
web_editor context automatically, making sure the method gets the
website_id all the time, removing the need to guess the current website
(now the get_current_website method uses the context website_id if any).
Note: other routes already have the website_id via request.env.context
if they correctly set website=True (this commit also adds website=True
for scss files customization routes).
|
|
* base, web_editor
Commit https://github.com/odoo/odoo/commit/44d659d0df6b50ae230af39a0d41a46b9549e2ae
made the system work with multi-website but fetching the assets content
was unfortunately not working in multi-website...
Also restore correct search operators.
|
|
website_id
Before this commit, the view create by editor (to change the scss file in asset for example) was not website specific.
Now we override (use the new hook method) to add the website_id.
|
|
With multi-websites, we expect qweb views to have a key set as it is the key
that is used to find duplicates (two views with same keys are duplicate, the
one with a website_id set is more specific than the one without a website_id)
Thus, 5ff87e803967 sort on key in order to find the most suitable one.
It would crash if a qweb view is created without a key (False) as it can't sort
`bool` and `str`.
This commit:
- Adds a generated key to scss views
- Adds an SQL constraint to avoid QWeb views without a `key`
- Generates a random key when creating a QWeb view without a `key`
- Handle False key in `filter_duplicate()` by extracting views with False key
before sorting, and then adding these views to the recordset
Note: We also want the key to be editable as it is now an important field with
multiwebsite. Thus, we removed the readonly on this field.
+ fix python tests as we now force key on qweb views with sql constraint
+ pep8
|
