FireFox extension: embed metadata into downloaded image

This will be a revisitation of an existing FireFox extension, you'll be tasked to fix a few bugs (of which I have the solutions most of) and to implement new features to existing code.

The extension does one thing really, that is to embed constant webpage data (URL, highlighted text and the Title of the webpage) into the IPTC metadata of a JPEG image (possibly a TIFF in this release too?) when downloaded to a users HDD. This allows for the image to be searched for easier and that data is useful for research with plenty of referential information.

This new version will provide compatibility with FireFox v3.5 (its v2 compliant right now) and to add new features like addressing scenarios where the image is not a JPEG; which will then be segregated into separate download folders for editing with 3rd party tools. Also translating the extension into six languages (I have the text, it just needs to be implemented).

Finally I will need to beta test it, time for that, and this will be a 1.0 release when finished.

## Deliverables

met.a v2 [4th revision - 1st August, 2009]

For the second phase of met.a I would like a more structured approach, with documentation accompanying each build - never attaining "v1.0" status until all sides are happy:?

(i) Fix old bugs

(ii) add NEW features

(iii) Testing time & Fix new bugs

-----(i)--FIX OLD BUGS--(i)-----

ia) Here the text selected seems to be held separate from the image window and will not embed:

[url removed, login to view]

ib) If downloading from here the image is downloaded as a program and appears corrupted (under Leopard 10.5.2, Intel Mac) but if downloading the larger image its fine. If an HTML page is downloaded (as seems to be the case) then put into "fodder" folder (see: ia pt.3) with accompanying text file for reference, unless this can be fixed:

[url removed, login to view]

ic) Downloading the larger img does not produce the save dialogue and does not download either:

[url removed, login to view]

id) Will met.a get the actual url from sites that force the same url across all pages? I think this effect is achieved by loading all pages in a frame, some dedicated tools even scramble the URL on purpose.. so I suppose not? Does met.a fetch the URL from the visible one in the address bar or does it look at the code, if so it might see the real one in these cases???

ie) feedback: "I liked the idea of what it is supposed to do.

It just doesn't seem to do it. Worse it disables the ability to save via the contextual menu.

OS X [url removed, login to view], G4, Firefox [url removed, login to view]

Not sure what other info to give you.

What is really odd is that once I selected disable yet hadn't restarted Firefox yet, then it worked?

I then attempted to download the source but it doesn't seem to be there.

The URL the firefox addon site listed was:

[url removed, login to view]

OK, here's the steps I did:

1) Went to the add-ons screen and selected disable for met.a

1a) Firefox tells me it won't be disabled until I restart Firefox but I choose to not restart yet.

2) Now try to use the contextual menu (that would be control click) on the picture and it works *and* there is meta data saved with the .jpg file.

3) Go "hmmm" and turn met.a back on in the add-ons screen.

3a) Firefox once more tells me about restarting and once more I ignore it ;)

4) Try to use a contextual menu (control click) and yup, doesn't work.

5) Turn it back off, don't restart Firefox and yet it works again.

What is odd is that it seems almost like it gets stuck, perhaps its a conflict with another contextual menu enabler I'm using because if I have meta.a on and then do the restart of Firefox I'm OK for awhile, I can control click on pictures to save and the meta data is stuck on the pictures. Then suddenly it stops.

I'm running OS X [url removed, login to view] on a PPC, so that would be a mix of Tiger and PPC (I'm not great with the Mac names but I think I got it right).

Thanks for the source but unfortunately the binary that's run to do the actual work doesn't have the source. But thanks.

I'm keeping it because it does work til it stops ;)

Then I just relaunch Firefox.

Brian"

if) Using the image name in the following case is better than the title of the webpage:?

[url removed, login to view]

fix with NEW feature (iia)

ig) The larger image will not be found and downloaded:

[url removed, login to view]

-----(ii)--NEW FEATURES--(ii)-----

KEY: ? (•) = Radio button

*drop FF2 support, concentrate on FF3.5*

iia) name: "Write Title:" "Webpage" (•) "Image" (•)

[default: Image] Preference pane = either but only one radio button

Amend existing "Write Title" feature to now allow the user to choose between the current "Image" setting (default) and "Webpage", which is the title of the Webpage that appears in the top of the window. The radio button should only allow the user to choose between "Webpage" (window title) and "Image" (image name given). For websites like [url removed, login to view] using the image name is best, but others not so.

ia) pt.1 name: "Recorded Delivery"

[default: on] Preference pane on/off option: checkbox

This option will cater for the formats that met.a can't, in effect all of the metadata that would otherwise be embedded is written into a text file (structure fig.a) and saved (with the same name as the image e.g. "[url removed, login to view]" would have a text file named "[url removed, login to view]") alongside the image in the default save folder. If turned off then files are left untagged and no text file are made.

The user will see each erroneous image and the accompanying text file directly beneath (if folder is sorted alphabetically) ready to make the decision to put them together themselves in a 3rd party app. The idea is that they will know what file(s) are not tagged and what info is relevant to those immediately without having to search in the browser history to chase up duff image downloads. This also applies to images that download as .html files (ref: bug section ib), unless a solution can be found, which would be better.?

When/if this does occur a small icon appears at the bottom toolbar of the browser to indicate that an image type(s) other than those taggable/managed with met.a were downloaded, a single click takes you to the containing folder (different if ia pt.3 is enabled). Once clicked this makes the icon (aka notification) at the bottom toolbar of the browser window disappear, until the next time met.a encounters an erroneous download.

? pt.2 name: "Fodder SFX"

? ? ? ? [default: off] Preference pane on/off option: checkbox

*available only if pt.1 is turned on*

An audible tone/motherboard beep sounded if image downloaded is other than accepted formats (JPEG/TIFF) and forwarded to the "fodder" folder (ia pt.3) (i.e. not taggable) - on OS X the tone to use is called: Submarine, on Windows it will be the default error sfx and I presume so on Linux.

pt.3 name: "Fodder folder"

[default: off] Preference pane on/off option: checkbox

Rather than keep offending files intermingled with positively tagged imagery have these (typically PNG/GIF/BMP) images shuttled to a separate folder called "Fodder" where each format has its own sub folder. (default location to desktop, but the user can decide an alternate location via "…" button next to title in met.a options).

fig.a: (text file structure for metadata entry on 'erroneous' image downloads)

-----------------------------------

Title: abc abc

Comment: abc abc

URL: abc abc

-----------------------------------

iib) name: "Format u?o??"

[default: on] Preference pane on/off option: checkbox

"sec [2]" user can change time for tooltip duration via drop down menu (no arrows) from 0.5, 1, 1.5, 2 or 3sec, default 2sec, located next to option title

When hovering over a non-taggable image (PNG/GIF/BMP) a sad faced pop-up bubble (embedded tooltip) will show, for two seconds (default), then disappear. The extension of the image dictates the graphic shown

iic) I can translate the extension into the following languages:

1. English (GB)

2. Spanish

3. French

4. Polish

5. Italian

6. Russian

The language in use is shown as a flag icon in the met.a options pane at the bottom right, the decision is not made by the user, its automatic. If not catered for the default to English.

iid) If the user only wants to download the thumbnail of the image shown then they must hold down the 'Shift' key and save as normal, the menu text would subtlety change from "Save Image As…" to "Save Thumbnail As…" (shifting met.a responsibilities)

iie) name: "Point-&-Click"

[default: off] Preference pane on/off option: checkbox + 'cmd'/'Ctrl' / 'Alt'

[default: off] SFX: checkbox (OS X=Tink/Win=beep)

For speed sakes downloading images will work with a single left mouse click while holding a hotkey, much like the Shift feature (ref: iid). This feature will have to override existing FF3 functionality for both images that are links and embedded images:?

Presently in Windows OS, the 'Ctrl' key + left mouseclick opens images that are *links* in a new window, while in OS X the 'cmd' key is used to the same effect. However when clicking links with 'Ctrl' in OS X a right-click menu pops up. But when clicking *embedded* images (i.e. not links) with the 'Ctrl' key under OS X a reduced save menu appears and using the 'cmd' key on an embedded image does nothing.

Using the 'Alt/option' key in OS X on a *link* will download it (to the specified download destination in FF's preferences), using it on an *embedded* image will highlight the image but do nothing.

So met.a has little option but to override existing functionality by defaulting to the 'cmd' key in OSX and 'Ctrl' key in Windows for such a key-combo. I base this on two reasons:?

1. That too many keys to remember is hard, while multiple text selection is done with these keys (in their respective OS environments) and so then using the same key in a single flow to download it (assuming text had been highlighted) is more intuitive (plus the key is bigger on a MacBook).

2. The feature to open in a new link is still a menu accessible option and *should* still work with normal text links.

This feature also has a secondary use as it will overcome sites which have the right-click menu disabled by default, normally preventing download ([url removed, login to view]). However in the event that this key-combo collides with an existing extension or preferred workflow, a user can either disable it (its off by default) or choose from the 'Ctrl/cmd' key or 'Alt' key in the options.?

~note that depending on the OS the optional keys will be displayed as 'cmd' under OS X and 'Ctrl' under Windows, never both if possible. The selection will be a radio button located to the right of the keys names, when clicked, greying out the other option (something better looking than a radio button if possible).

n.b. This feature must also to work with any highlighted text (ref: (iih)).

When downloading from the menu a user has confirmation that the process worked because the menu disappears, with the key-combo an audible sfx will be an optional method of feedback from met.a that the process worked. This being (optional check-box): Mac OS X = Tink / Windows = beep titled "SFX". Otherwise a users could be left unsure.

iif) name: "I'm Lazy"

[default: off] Preference pane on/off option:?

This will *always* embed all visible text from a page into the downloaded image, thus avoiding the need to select any text. Useful for projects where you will always need all of the data on a page. Or, if you're just lazy.

.nb The TIFF format file size maxes out at 2GB, while JPEG is unique in that the comment field can handle 64kB.

iig) When selecting text that has a link in it, convert the text to show the link in the metadata e.g. "home" (hyperlink) changes to "home [[url removed, login to view]]"

iih) Make sure that the new feature from Mozilla in FF3; which allows multiple portions of text to be selected from a page (holding 'Ctrl' (Win) or 'cmd' (Mac) and dragging to select), is compatible with the feature set of met.a.?

**This would be really nice to use in conjunction with Ctrl/cmd and left mouse click to download images via the "Point-&-Click" feature (ref: (iie)).**

iii) Extend meta tagging to the TIFF format.

iij) Trimming the fat off the extension in terms of size and reduce memory footprint is always something to bear in mind, especially with more features being added. Ultimately met.a is a background app and must not affect the users workflow so speed is important. But I know you know that.

-----(iii)--TESTING TIME / FIX NEW BUGS--(iii)-----

Fix the inevitable bugs that result from the new features via feedback from users I will give the extension to for BETA testing during the forming of the extension:?

FireFox v3 (the FF3.5 release should work well with extensions made now that are within guidelines released from Mozilla)

PPC/Intel Mac: Tiger OS/Leopard OS/Snow Leopard OS?

Intel Windows Vista OS

Linux OS

UNIX OS

----

Other requirements:

a. Source to be made available via Google Source code ([url removed, login to view]), although I want met.a to be nimble for source code usefulness it would be nice to have lots of comments, I think you said before you do this already. I don't know f this means having two parallel version of met.a as stripping out comments might make it faster? I will need the code handed to me separately for the online upload as I found it problematic the time before getting to it I think, thanks.

b. Make sure FF sees new version for future updates, as you say it should care of that anyway.

c. Incremental BETA releases need to be supplied with new feature(s)/bug squash documentation?

d. The layout or the menu will be thus:

met.a me more of less

[x] Write Title: (•) Webpage ? (•) Image [default: on]

[x] Write URL [default: on]

[x] Disable highlight embedding ? [default: off]

-------------------------------------------------

[x] Recorded Delivery: [default: on]

(•) Fodder SFX [default: off]

(•) Fodder folder […] [default: off]?

-------------------------------------------------

[x] Format u?o??: ? sec [2] [default: on]

-------------------------------------------------

[x] Point-&-Click: [default: off]

(•) [cmd (Mac)/ctrl(Win)]

(•) [alt]

[x] SFX [default: off]

-------------------------------------------------

[x] I'm Lazy [default: off]

*country flag* [GB]

---about me page---

add homepage: [url removed, login to view]

add my email to about page credit

add language translators to credits "Champinion, Lorenzo, Sara, Olga"

Ingeniería Linux SO Mac Microsoft MySQL PHP Gestión de proyectos Arquitectura de software Verificación de software Windows Desktop

Nº del proyecto: #2817924

Sobre el proyecto