- browser extension: reads datamatrix code (not other kind of codes) from camera and either link to an already existing sample entry or create a sample in an inventory and paste the link into an text field in rspace ELN (support for eLabFTW ELN is work in progress, check out the Roadmap)
- Hint: skip reading and check out the step-by-step guide or the Schritt-für-Schritt Anleitung
- open a basic electronic notebook document
- open the edit mode for an iframe (i.e. regular custom text field)
- set the text cursor where to insert the link(s) (blinking text cursor)
- click the "link sample" button (top-left corner), the browser camera should open in transparent overlay in front of the ELN (camera connection maybe have to be allowed once)
- present the code (any code supported from html5-qrcode) in front of the camera till it recognize the code (size matters!)
- if a green text field pops up after decoding this code have been found in the database and the results will be shown below
- the "thumb up - insert button" insert these search results as link(s) to the document
- if a yellow text field pops up after decoding, this code couldn't be found in the database and can be used to create a new sample entry connecting the current ELN document and the physical sample connected to the code
- a custom sample name for the newly created sample can be insert into the yellow dotted line text field below
- or use the already prefilled default sample name
- the "thumb up - insert button" create a new sample entry and insert a link to the sample created into the document
- Python >= 3.10
- Flask >= 2.2.2
- Flask-Cors >= 3.0.10
- requests >= 2.31.0
- fill the api_secrets.example file and rename it to just ".json" api_secrets.example
- changed to the current state of the art manifest version 3
- package the extension as .crx file (now runable without developer mode, e.g. allow for automatic distribution)
- sucessfully tested on android 12 tablet, windows 10, windows 11 surface
- implemented the flask backend as productive use server
-
user experience made more interactive and flexible
- change from "click to scan code" to "continuous scan until a code is found" (a blue "wandering bar" indicates the scan process, it should improve the user experience that the scan works)
- new feature: search for a previously created sample with direct link to the search result
- new feature: insert a custom sample name right at the time of creating a new sample entry
- the default samplename is now additionally stored in an extra field (defaultELNconnection) to protect the user from losing the important information about the connection of the ELN document to the sample tracking system
- i.e. default sample name: Qrcode-decodedText_ELN-documentName_ELN-document-uniqueID_ELN-document-creator-username_timestamp-at-creation
- Remember the basic idea of sample tracking! "each sample must be linked to a protocol describing its creation"
- two new buttons only visible after sucessful decoding: insert button (insert link(s) to current text cursor) & abort button (close camera, abort sample creation, delete search results)
- created a new coloured message field at the top to distinguish more quickly between the three possible search results:
- green (the decoded code has already been entered in the database, check out the found entries listed below)
- yellow (the decoded code could not be found in the database, an entry for it can be created after entering a sample name)
- red (TODO: some error messages)
-
switch from backend to frontend decoding
- previously backend decoding (pylibdmtx) was used now the more common frontend decoding (javascript based in-browser) with HTML5qrcode (html5-qrcode) is used,
- BUT this solution is often slower (because of worse segmentation algorithm), have issues with smaller codes and some "unusal structured" codes (see TODO Roadmap)
- since HTML5Qrcode support a lot of different codes it now can use mostly every code exist (i.e. QRcode, barcode, datamatrix etc.)
- no more privacy issues as no image is stored or leaves the local computer
-
added a german translation of the "step-by-step tutorial"
-
a lot of refactoring, restructuring of the sourcecode and changes of the roadmap priorities
- use greyscale images to decrease the imagesize send to the server, it also speeds up the decoding
- bugfix: deactivate camera after click on "link sample" button and after insert a link to the ELN (no matter independend of sucess)
- known Bug: impossible to insert multiple links into an empty field without saving between inserts
- bugfix: it now should read the cursor position (the correct cursor position should now be found and the link in the exact place (before and after blank characters are inserted & the link is always in its own line, because it is very long)
- updated roadmap
- first full working version (public release)
- known bug: it doesn't read the cursor position precise enough, so better to use at the end or the beginning of a field
- known bug: if a datamatrix code is being used more then once it only paste the link to the first entry found (what should happen if there's more then one entry connected to one datamatrix code? paste all codes? even if there're plenty?)
List in no particular order ;) (Priority [1..5])
- [5]: add full eLabFTW support
- [5]: improve the scanspeed (e.g. shearing correction, image improvements)
- [5]: show an error texts if something unusal happen (especially a warning if the a user tried to add a link without being inside an iframe; connection issues)
- [4]: add useability of physical barcodescanner devices
- [3]: add more customizations (edit the link text, insert eLabFTW API calls)
- [3]: prevent barcode entries from being deleted
- [3]: add button for switching cameras (for devices with front and back cameras, i.e. tablets or smartphones)
- [3]: improved comprehensibility of the interface (hover infos, more intuitve button images)
- [3]: speed up the decoding, it has not really gotten better with frontend decoding
- [2]: improve folder structure/ refactoring (especially use full HTML5 capacities not only javascript and CSS)
- [2]: add batch-scan possibility (depends on user feedback, so this possible function will be re-evaluated after enduser-tests)
- [2]: if multiple samples are found for a code, allow only one of the samples to be linked into the text (conceptual question if such a function should be allowed for sample tracking, as a code should only be used for one sample object; there might be scenarios where this makes sense, on the other hand a double use of a code indicates an error in the database; will be discussed after user feedback)
- [2]: create a mode to search for samples anywhere not only inside a document
- [2]: package the extension to use it as regular extension
(TODO: create a mode to search for samples anywhere not only inside a document)
- [5]: speed up the decoding
- [5]: change the workflow so it scans for codes all the time and only close and insert after it found something (no more record button pressing)
- [5]: show an error texts if something unusal happen (especially a warning if the a user tried to add a link without being inside an iframe; no active iframe = nothing can be paste into an active iframe)
- [5]: use the userName transmitted from the client to create samples not the API username (in the final version samples will be created by the admin, so the created username should be the one who initiate the creation of the sample)
- [5]: add uniqueID of the current ELN document
- [4]: add a button to directly name the sample from extension interface
- [4]: show frame sent to server after pushing record button instead of the video stream (maybe show a frame surrounding the decoded matrix code)
- [4]: add customization support (edit the link text, insert eLabFTW API calls)
- [4]: create a final server structure to be used instead of "non SSL flask development server"
- [3]: improved comprehensibility of the interface (hover infos, open-close camera button, easier to understand record button)
- [3]: add eLabFTW support (full support)
- [3]: add an option to not use the "record button" to link a sample (instead scan every second and insert link after sucessful decoding)
- [2]: improve folder structure/ refactoring
- [2]: include another decoder to also decode other kind of codes like QR codes and barcodes
- [2]: add batch-scan possibility (function depends on user feedback)
- [2]: package the extension to use it as regular extension
- [1]: add button for switching cameras (for devices with front and back cameras, i.e. tablets or smartphones)
- [1]: show a progress/ waiting bar