87 replies to this topic

[dafi]

This week I reached my 62nd extension localized with the precious help of this extension. Just to let you know how efficient and useful it is.

Sound good for my ears
When we (you) reach 100 extension we can organize a party like this.
Obviously this is a joke

I have noticed there are still sometimes double similar entities created with same name and same strings, I don't know why because generally there is the appending number that allows entity names to be different, but sometimes there is no such number.

Bad news, may you send me more info so I try to reproduce the problem.
I can't live knowing my friend Goofy encounters similar bugs when he works

In a distant future, it would be good to have this feature :
For xml files, the user should have the choice between
- automatic detecting and replacement (just as it works fine now)
- selection of items to be transferred just as it works fine for .js files now

I'll add these enough simple modification as soon as possibile, please apologize me for late in fix these problems.

Thank you again, dear Davide, for the very kind and smart work you have done !

My pleasure, I'm happy when people use my bonsai apps

[Goofy]

may you send me more info so I try to reproduce the problem.

As a sample to give a look, here are 2 files, the dtd is obviously the output result for the xul one.
(Some entities are same with same entity names)

Thanks for your patience

[EDIT] ouch my bad. I realize now that the problem (as in zipped files) occurs because I set the max char to 30 and the strings where <30.
I just tested again with char max set to 20 only and all entities have appended numbers so there are no more similar entities, though strings are similar

Attached Files

•  two_files_with_duplicate_entities.zip   1.44KB   22 downloads

[Goofy]

hello Davide. I don't know what stupid mistake I am probably doing and i would like you to giev a look.

I am currently "extenalizing" an extension which includes accesskeys and though I duly listed "accesskey" in my attrb.txt file, the process seems to ignore accesskey.

here I attach my txt file, and one original xul file + its externalize version skipping the ak
I am (almost) sure the accesskeys were detected in previous versions...

Thanks for any clue you could give

Attached Files

•  attrs.txt   158bytes   19 downloads
•  distrust.zip   972bytes   24 downloads

[Pedro]

It is quite off topic, but I havent caption in my attrs.txt file... is my old file?

[dafi]

hello Davide. I don't know what stupid mistake I am probably doing and i would like you to giev a look.

I am currently "extenalizing" an extension which includes accesskeys and though I duly listed "accesskey" in my attrb.txt file, the process seems to ignore accesskey.

here I attach my txt file, and one original xul file + its externalize version skipping the ak
I am (almost) sure the accesskeys were detected in previous versions...

Thanks for any clue you could give

I'm investigating...

It is quite off topic, but I havent caption in my attrs.txt file... is my file old?

Consider actually don't exist official files, so it's possibile you have a different (not necessary old) file

[Pedro]

Thanks Davide

I try and I have the same result of Jean Bernard

I am now externalizing an extension called Hyperbookmark, and I notice that the entity with only one letter aren't processed... may this help you in your investigation?

[dafi]

hello Davide. I don't know what stupid mistake I am probably doing and i would like you to giev a look.

I am currently "extenalizing" an extension which includes accesskeys and though I duly listed "accesskey" in my attrb.txt file, the process seems to ignore accesskey.

here I attach my txt file, and one original xul file + its externalize version skipping the ak
I am (almost) sure the accesskeys were detected in previous versions...

Thanks for any clue you could give

Found the problem
The new "Regular Expressions used to exclude values" skips the string like "u" or "T".
To verify this simply create a file containing the string "nonenone" and use it and all value are generated.
This is a very invasive bug

[Goofy]

Thanks Davide

It is quite off topic, but I havent caption in my attrs.txt file... is my file old old file?

Davide is right, this is just one of the files I am currently using, adapting it to the extension I am dealing with (add/remove items as you wish). As an example, today I encountered "tooltipText" so I added it (ah dev have their fancies and whims )

[Pedro]

Thanks Jean Bernard... now the matter is clearer

[dafi]

hello Davide. I don't know what stupid mistake I am probably doing and i would like you to giev a look.

I am currently "extenalizing" an extension which includes accesskeys and though I duly listed "accesskey" in my attrb.txt file, the process seems to ignore accesskey.

here I attach my txt file, and one original xul file + its externalize version skipping the ak
I am (almost) sure the accesskeys were detected in previous versions...

Thanks for any clue you could give

Try the file in attach with a modified regular expression.[attachmentid=1963]
This is simply a workaround (may not work) that skips string values with length greater that 2 chars instead of 1 char.

Attached Files

•  re.txt   36bytes   31 downloads

[Goofy]

Thanks Davide, I will try tonight.
Now I am skeptical about your answer : the problem is not to skip values > 2 chars, but to detect one-char values like "A" when accesskey="A".
But I am so ignorant about regexps that I probably misunderstood the whole trick (?)

[Goofy]

Thanks Davide, I will try tonight.

Done: it works fine with your workaround. Thanks!

[dafi]

Done: it works fine with your workaround. Thanks!

Honestly I know this is only a workaround not a solution.
It is the quickest way to continue to use Externalize without waiting for a new build.
Maybe a solution should allow to choose the technique for discarding values per tag.
For example

CODE

  Tag/Attribute         Discard value RegExpr
     accesskey               None
     label                  RE1, RE2

Imagine a UI to obtain a similar result.
Adding an UI to handle files is a priority and I'll add it ASAP (but not before December )

Stay tuned

[Goofy]

Maybe a solution should allow to choose the technique for discarding values per tag.
For example

CODE

  Tag/Attribute         Discard value RegExpr
     accesskey               None
     label                  RE1, RE2

Imagine a UI to obtain a similar result.
Adding an UI to handle files is a priority and I'll add it ASAP (but not before December )

Stay tuned

I am impressed and not sure to understand how the per-tag regexp* will work , but I am ready for Goofytesting even if I must wait for Santa Claus

*For one thing I don't know about regexp syntax (I have read tutorials but I have no mathematician brain- well I suspect I have no brain at all to tell the truth) but what makes things more complicated is that every file may need a specific choice, and also in the same file sometimes a string must be externalized or not.

For example, it often happens to have in the same file
value="Davide is Great" (-->must be translated in every possible language)
and
value="Codestuff" (--> must be left as it is to make the extension run smoothly)

So the workaround here should be to allow user to select step by step what must be externalized and what mustn't be

[Goofy]

@Pedro for information, I am currently trying to externalize Chipmark extension

[Pedro]

@Pedro for information, I am currently trying to externalize Chipmark extension

Thanks for the info Jean Bernard

[Goofy]

[Done]

[dafi]

Finally I've restart to develop and I've added to externalize the ability to select items to externalize also for XML files.

The version 0.0.5 is here.

Please install it on a clean profile, I've dramatically changed the code so some regression is possibile.
The update system is disabled because 0.0.5 is considered a beta

I wait for your hints...

please post here
- bug report
- regression report
- improvement request
- new features request

This new version contains the base code for version 1.0

PS
The line number isn't present for xml files, this is due to actual parsing algorithm so don't worry about it

[Goofy]

Dear Davide it is just great !

apart from this little displaying annoyance for the extraction dialog
[attachmentid=2231]
(appears at first launch, it is fixed once it is manually resized)

I cannot find anything wrong with this version. It is very handy to manage possible automatic entity duplicate names and of course to tell precisely which string should be externalize or not

Attached Files

•  externalizeDialogSize.png   68.31KB   28 downloads

[Goofy]

Ooops I found a disturbing issue...

I have seen that the original strings are removed from one xul file BUT NOT replaced with the newly created entities. More surprising, the new entity is duly reported in the dtd file with the corresponding strings

Here I attach the original xul file, then the processed one (with missing values and labels!) and the resulting dtd.
[attachmentid=2237]
[edit] I think the issue occurs each time I change the entity name in the editing dialog before launching the process (I will check that)

[edit2] I checked and unfortunately it trims again some labels and values even if I did not change the entities' names

[edit3] for information the same test file is correctly externalized with the previous version, so it is obviously an issue related to the new code part.

Attached Files

•  TestFiles.zip   1.62KB   17 downloads