Python3 + appscript (appscript.sourceforge.net), which I wrote after getting sick of dealing with AppleScript. Basically what JXA should’ve been (in addition to Apple’s answer to Node), had Apple’s Mac Automation team been run by a competent manager, but it failed so hard Apple scrapped that entire department in 2016. Whatever the future of Apple automation may be, it lies with the Siri/Shortcuts team now. ... View more

Alas, Adobe’s Creative Cloud app insists my installed copy of AI 25.0 is the latest version available to me, so I’m unable to upgrade to test it myself.   1. The most obvious solution, assuming this is in-house only, is to revert to an earlier version of AI that you know works (the CC app allows you to install multiple/older versions), file a bug ticket with Adobe describing previous and current behaviors, with test scripts/screenshots/error messages/etc to help them understand the issue, and hope they fix it before your next upgrade.   2. Still working with the problem “do javascript” command: if it won’t take a string containing JavaScript code, try writing that code to a [temporary?] file and try to pass references to that in as many different forms as you can think of to see if one of them works: POSIX path string, AppleScript “alias”, “POSIX file”, HFS path string (ugh), “file”. “Poke it all over with a stick until it does what you want” is, depressingly, one of the more effective troubleshooting techniques for AppleScript.   3. If you’re looking for an alternative anyway, the AppleScript interface is not the only way in. Depends on what you’re trying to achieve and how far you’re willing to go to achieve it. There’s the C++ SDK, of course, but I doubt you want to deal with that. Perhaps the CEP/Adobe Bridge users here can provide some suggestions? (I haven’t looked at how Bridge itself communicates with other Adobe apps.)   I do remember one enterprising AI engineer stuffing a Node-based webserver into a CEP panel:   medium.com/adobetech/how-to-build-a-node-js-server-in-a-panel-ba1d63ea67e2   (Needless to say there are security issues to consider when running any sort of webservice on a desktop; limiting client connections to localhost being an obvious start.)   And once UXP does ship for AI, who knows what options will be available then? Though it’ll be another year (or two, or three?) before we get that. ... View more

Restart your computer, reset Automation security settings, and run your scripts afresh is good basic troubleshooting advice, but it doesn’t sound like that’s the cause.   I’ve not touched AI’s AppleScript interface in over a year (recent projects have all been in JSX), so haven’t been following exactly what’s changed and when, but this may be related (“foo” is me):   stackoverflow.com/a/58856316/7926970   Basically, the Illustrator application dates back to Mac OS 7 days, and some parts of its guts use pre-OSX, pre-sandboxing, macOS-defined datatypes and APIs (e.g. Carbon Alias and File Managers) that Apple has long since deprecated and will eventually phase out. Therefore Adobe needs to replace old AI code that uses those old, dying APIs with new code that uses modern, supported macOS APIs. Unfortunately, AppleScript, being another System 7-era technology and long neglected by Apple, is still using some of those old dying datatypes itself, and while it supports some (though not all) of the newer types this support is kludgy, inconsistent, worryingly brittle, appallingly documented. Plus, half of these datatypes don’t play well with the other half.   Now add some current AI engineers, who probably have little-to-no familiarity with AppleScript or Illustrator’s support for it, typing to upgrade neglected quarter-century old code so that it works well with modern macOS sandboxing, and a small but ardent base of heavy AppleScript users who rely on this whole house of cards to keep on working while all this chop and change is going on.   I did observe at the time that while the old AppleScript “file specifier” type still in use in some parts of AI (e.g. “file path” properties) was horrendously obsolete and long overdue for replacement, but using plain old strings to describe file system paths was a really bad idea, precisely because there’s no 100% reliable way to distinguish a string specifically describing a file path from a string that contains any old text. (This is the whole reason for having different datatypes in the first place.) Plus, the way that AppleScript performs string↔︎file type coercions (badly, inconsistently, and non-reversibly) means that older AppleScripts that used “file-like” objects would break where path strings are now expected, and vice-versa.   The appropriate change I said at the time would be for these properties to use “POSIX file” («class furl»), since that makes it clear what type of data it is, even in a command such as “do javascript” that accepts either a value that points to a JavaScript file or a string containing JavaScript code. So it may be that later 24.x releases and 25.0 changed again from using path strings to another “file-like” type, but with all these rolling changes—and with AppleScript’s own support and documentation being horribly confusing—it wouldn’t surprise me if it all gets tangled up in itself, particularly if AI is also trying to be “helpful” by still allowing scripts to pass file paths as plain strings. Which, as I say, is liable to turn into an error-prone mess for any command that accepts code files and code strings interchangeably.   To be clear, even if the above is the cause of the problem (which it may or may not be), this is Apple’s fault for running AppleScript into the ground over the last 15 years, not Adobe’s for trying (and failing) to make the resulting confusion work. Honestly, I’ve built the most advanced Illustrator automation on the planet, and I did using AI’s AppleScript interface (though not AppleScript itself) because when it works it is incredibly powerful, elegant, and a pleasure to work with (much nicer than JSX!). And even I’m counting the days till Adobe ships UXP for Illustrator now, as AppleScript infrastructure looks increasingly like a bitrotten liability, not a platform on which to build a future.   --   TL;DR: Working with files in AppleScript is a nightmare, period, and not getting any better. ... View more

One thing: check the style names used in your script exactly match the style names in the AI document. AI can be a bit persnickety about differences in case. ... View more

Works here: tell application "Adobe Illustrator" -- 2021 tell document 1 apply character style character style "Test" to text range of text frame 1 end tell end tell     ... View more

Looks fine in AI 2021 (Mac, but Win should be same). What version of AI are you on? ME support was pretty nasty in older versions, so upgrade if you’ve not already done so. ... View more

So go ask your customer which barcode format they need. It’s ridiculous for them to ask you to produce “5-digit barcodes” but not tell you “EAN-13” or “UPC-12” or “Code-39” or “Code-128” or “Pharmacode” or whatever. Because all those different barcode formats will encode those 5 digits in very different ways. There’s a huge range of barcode formats, all used for different jobs, countries, and markets, and if you don’t know precisely which type of barcode to produce then you’ll most likely produce the wrong sort. Which, if you’re lucky, their barcode readers will reject as “unreadable”; but at worst might read as a completely different number—and then you’re really in trouble.   Honestly, generating barcodes is one of those jobs that if you don’t get nightmares about What Might Go Wrong, you really shouldn’t be doing it. Because you don’t understand it well enough to know if you’re doing it right.   Here’s a link to the BWIPP site that lists all the different barcode formats that the BWIPP library supports:   github.com/bwipp/postscriptbarcode/wiki/Symbologies-Reference   BWIPP supports a wide range of barcode formats, both common and rare, so it’s very likely the format your customer needs is somewhere on that list, but they need to tell you which one it is, not leave you to guess.   ... View more

That is not an EAN-5. This is an EAN-5, generated with BWIPP: (45454) (includetext) /ean5 /uk.co.terryburton.bwipp findresource exec   And, as I say, EAN-5 is only used as an adjunct to EAN-13; it’s not found on its own.   Yours might be a Code 39 or some other variable-length type, but I am not going to guess. If you can’t tell us what barcode type it is or what you’re using it for, we cannot advise you. The last thing you ever want to do is to produce an incorrect barcode. That could get you in a lot of trouble. Global product recalls happen for less.   I strongly recommend 1. you read up on common barcode standards and symbologies so you understand what you’re dealing with, and 2. you ask whoever you’re producing these barcodes for to tell you what type of barcode they require, and then ask them for whatever additional information (scale, guards, BWR, font, etc) you need to produce the barcode artwork. Then you can go find the appropriate tool to generate it. ... View more

Rather than asking how to generate a “5-digit barcode”, it would be better to explain what you’re creating it for and show some sample artwork   For instance, if you’re making book covers then I’m guessing you want an EAN-13 with supplementary EAN-5 code. (The BWIPP library has an option for producing those, as will more advanced barcode generation apps.)   Or, if you want to encode arbitrary numbers for some other purpose, there’s a variety of barcode types to choose from depending on what they’re being used for (e.g. shipping/equipment labels) and how they’ll be read (point of sale, warehouse scanner, smartphone camera, etc). ... View more

OP might be referring to the supplementary EAN-5 code that sometimes appears after an EAN-13 on books.   en.wikipedia.org/wiki/EAN-5   For barcode generation I have in the past used the free open-source BWIPP (github.com/bwipp/postscriptbarcode/wiki) which supports an impressively wide range of 1D and 2D barcode types. Catch: it’s a PostScript library, so you have to write a tiny bit of PS code in order to use it.   (I also don’t think its barcodes are as pretty as, say, Esko’s. (But hey, they’re barcodes. If they scan, they scan.)   The BWIPP wiki documentation lists the currently supported barcode types, the data they accept, and the standard- and barcode-specific options that can be used, along with plenty of examples of use.   FWIW I typically top and tail the original barcode.ps file with: %!PS-Adobe-3.1 EPSF-3.0 %%BoundingBox: 0 0 1440 1440 and: gsave SCALE SCALE scale 72 72 moveto BARCODEDATA OPTIONS /BARCODETYPE /uk.co.terryburton.bwipp findresource exec grestore showpage  replacing the SCALE, BARCODEDATA, OPTIONS, and BARCODETYPE placeholders with actual values, then resave it as an .eps file which AI can import via File > Place†. (The BoundingBox line is only there to convert the PS file into a valid EPS; just make it sufficiently large that it won’t crop the barcode at normal scales.)   Plenty other free/paid options too (some of which might even use BWIPP under the hood), although I agree OP should do some background reading first, if only so he can better express his requirements here.   --   † Assuming the syntax and arguments are all valid. (If anything’s invalid you just get a big fat nothing; neither AI nor BWIPP provide helpful information when PS errors occur.) ... View more

Sounds automatable. Could you post some annotated screenshots to clarify your process? ... View more

Are you talking about a script that already exists, or are you looking for someone to write you one? ... View more

Does Vassily’s reply solve your problem?   Imposition (laying out multiple artworks to be printed all at once on a larger size of paper) is traditionally handler by the printing vendor, along with cutting/folding/binding the printed results. If you’re sending the artwork files to an outside printer, talk to them first.   Or are you printing them yourself on Avery-style labels, in which case you’ll want to lay them out to match the pre-cut sheets? If that’s the case, and the individual artworks are already generated using sequential file names, I would suggest importing those files in batches into an Illustrator/InDesign template and print from there. (You can often find free templates online, although they’re more commonly for Word.) ... View more

Illustrator already has basic data merge support which you should investigate first.   Here are a couple (free/paid) tutorial videos:   https://www.youtube.com/watch?v=mpwXj2E7YNA   https://www.lynda.com/Illustrator-tutorials/Welcome/450909/489328-4.html   Or, if you prefer text-based guides:   https://helpx.adobe.com/illustrator/using/data-driven-graphics-templates-variables.html   https://www.christianda.com/blog-variable-data.html   If you require a more advanced (scripted) solution, post your detailed requirements to get further advice. ... View more

To be clear, JSXBIN and friends are source code obfuscators, not compilers. What they do is make it harder for others to decipher your code by making it difficult to read: removing all comments and whitespace, replacing your descriptive variable and function names with randomly generated ones, add fake code that never actually runs but helps disguise the real code under reams of noise, etc. That’s about as good as you can hope for with a language such as JavaScript, where all programs are loaded and run from plain text source code files.   Some scripting languages (e.g. AppleScript) can save scripts in binary (bytecode) format, which requires specialist knowledge to convert back to human-readable code, and some (e.g. Python) have tools that can convert your scripts to C code and compile that, which would take a very skilled programmer a lot of time to reverse. Or you’ve always got the option of implementing your product as a C++ plugin, which again is a lot more expensive for others to reverse-engineer—and a lot more expensive for you to write in the first place.   That said, even programs written in fully compiled languages such as C++ and distributed in binary form only (e.g. Adobe Illustrator) aren’t immune to reverse engineering given sufficiently skilled and motivated programmers of dubious morals. (Hence all the w*r*z sites offering “cracked” versions of commercial software.)   It’s all a matter of weighing costs vs benefits. Your hope is that the time and labor (and possibly legal) costs to them of hacking your code significantly outweigh any benefits they might hope to gain, and so deter them from doing so. If it’s a simple low-cost script you wrote for a handful of clients then running through an obfuscator is likely sufficient to put them off trying to extract your “secret formula” and just pay their fair dues to use it. OTOH, if it’s a major software product that’s going to retail globally and generate millions of dollars a year then, yeah, someone’s likely going to crack that code. But that’s why you have copyright, software licenses, legally-binding contracts, and your own fleet of well-paid lawyers so you can go after those violators with the law.   HTH   --   p.s. Encryption isn’t really a practical solution as your customers’ computers need to be able to run your software themselves—which they can’t do if it’s been locked with encryption and only you have the key. So even if you do encrypt your main code, you still have to supply a way to decrypt it along with the key for performing that decryption… and you’re right back to where you were to begin with. Again, if you’re thinking on those lines you’ll be much better investing in a big bucket of chum for your legal department.   p.p.s. Were you doing high-end InDesign automation, you’d have another option which is to offer your automation as a hosted service (aka “SaaS”) where users upload their source files to your website which automates them using InDesign Server and returns them the finished files. That way users never have access to your proprietary automation code so cannot steal it (short of breaking into your server and/or data center and committing outright theft). But of course you have to consider if that operating model works for both them and you (and unless it’s earning you six-figure sums the costs of licensing IDS alone mean it’s not worth your trouble). That’s not an option with Illustrator though, which only allows existing license-holders within your own organization to use it. ... View more

1. File paths must be absolute, e.g. "/Users/carl/testFile.pdf"   2. If you’re on AI 2020, the file path must be given as a string:         set imageFileName to "/Users/carl/testFile.pdf"    For earlier versions, it must be a file object:         set imageFileName to POSIX file "/Users/carl/testFile.pdf"   3. To view an application’s AppleScript dictionary in Script Editor, choose File > Open Dictionary. PDF-based documentation  (same info, but also includes example scripts) is also available at: www.adobe.com/devnet/illustrator/scripting.html ... View more

Hi Sharon,   I’ve just emailed you regarding automating your tidal charts—which are cracking, by the way.   This is the best forum for Illustrator-specific scripting questions. It’s also a good first stop if you’re looking for freelancers to do work for you, in addition to more general freelance sites such as Upwork.   If you wish to learn AppleScript/JavaScript/VisualBasic then there are other forums that can help with that, and I’m sure folks here can recommend a good beginner-friendly book to get you started. (Disclosure: I lead-authored Apress’s Learn AppleScript, 3rd Edition.)   HTH ... View more

If you succeed in negotiating public server licensing terms with Adobe, come back and let us know.   Failing that, your choices are:   1. Write a desktop- or CEP-based tool that graphics suppliers can use to open the file in Illustrator, automatically extract all of the pertinent information (text frame content and coordinates) via AI’s scripting/C++ API, and uploads that data and the [PDF-compatible] AI file to your online review app to process and display to others.   2. Upload the [PDF-compatible] AI file to your online review app and attempt to extract all the text from the raw PDF data using 3rd-party PDF-reading software (a fairly horrible and not wholly reliable black art).   One other tip: If the original pack copy is available from a third-party data source (e.g. MIS system, Excel spreadsheet), you might consider reconciling that structured source text with the unorganized final text taken from the artwork. Pack copy documents (assuming a consistent machine-readable layout) will contain additional useful information—e.g. field names, glossary texts—which you can use to better identify and organize the extracted text for improved presentation and semi-automatic checks. Also good for detecting text amends and other alterations made to the artwork which may differ to the original pack copy, and flagging those for further attention. ... View more

Check the file path is correct. ... View more

Try: set jsFile to POSIX file "/Applications/Adobe Illustrator CC 2018/Presets/en_US/Scripts/Submit_Files.jsx" tell application "Adobe Illustrator"    do javascript jsFile end tell ... View more

Not something I use. AppleScript’s the obvious choice given KM’s support for it, although there are also folks doing work with Adobe’s CEP if you can tie KM into Node. ... View more

I wouldn’t waste time on JXA. It’s flawed, buggy, and effectively unsupported (it bombed so hard, Apple disbanded the team responsible). It works up to a point, but try doing something non-trivial and it breaks, you’re stuffed. Same goes for Apple’s Scripting Bridge before it. Doubly so when dealing with older Carbon-based apps. You’re best sticking to Adobe’s built-in JavaScript and/or AppleScript; neither is perfect, but at least they work and have documentation and user communities worth a damn. If you _really_ want to use JavaScript + Apple events then node.js + nodeautomation should work okay, but caveat emptor as I don’t provide support. ... View more

@jlockman: While I can’t speak for others here, our own customers’ (mid-size GSPs) concern is a simple one: each time an operator starts automation on her own desktop it effectively takes over her machine, preventing her doing any work until the automation has completed. Considering that  a typical  operator is 100x more expensive than a typical CC application (for which they already have site licensing), offloading that processing work to another machine on the same intranet would eliminate this costly waste of human time. Furthermore, the EULA does indicate this kind of restricted internal use is allowed for at least some Adobe applications. From the Illustrator CC 2018 EULA: 2.1.7 Server Use. 2.1.7.1 Except as otherwise permitted in the Documentation and subject to license restrictions stated in this agreement, Customer shall not install the Software on a Server. If permitted in such Documentation, then Customer may install the Software on a Server for the purpose of allowing an individual from a Computer within the same Internal Network (“Network User”) to access and use the Software. If expressly permitted by Adobe, “Internal Network” may include web hosting services with dedicated physical server space and restricted access to only Customer. The Network User who has access to or may use such Software on the Server is referred to as “Server Software User”. Neither the total number of Server Software Users (not the concurrent number of users) nor the total number of Computers capable of accessing the Software installed on the Server may exceed the Permitted Number. By way of example, if Customer has purchased 10 Software licenses (Permitted Number is 10) and Customer elects to install the Software on a Server, then Customer can only allow up to 10 Server Software Users access to the Software (even though Customer may have more than 10 Network Users or fewer than 10 concurrent users of the Software). 2.1.7.2 For clarification and without limitation, the foregoing does not permit Customer to install or access (either directly or through commands, data, or instructions) the Software: (a) from or to a Computer not part of Customer’s Internal Network; (b) for enabling web hosted workgroups or web hosted services available to the public; (c) by any individual or entity to use, download, copy, or otherwise benefit from the functionality of the Software unless licensed to do so by Adobe; (d) as a component of a system, workflow or service accessible by more than the Permitted Number of users; or (e) for operations not initiated by an individual user (e.g., automated server processing). The key clause (my bold) states that “if permitted in such Documentation” installing software on an intranet-only server for use within the organization is allowed, as long as each user already has a desktop license. The question is: to exactly which documentation do we refer in order to determine whether or not intranet-server use is permitted for a given CC application? It would be most helpful if Adobe’s legal department could resolve the confusion here, both amongst customers and its own support staff. ... View more

Why do you want to use UDP? It's an unreliable transport mechanism, intended for stuff like streaming video where transmission errors are acceptable. You don't say what OS[es] your app runs on or the complexity of the operations you want to perform, but here's a couple options: 1. Use the OS's native IPC mechanism (Apple events on Mac, COM on Windows) to talk to Illustrator. You'll find the Illustrator API docs at https://console.adobe.io/downloads/ai under 'Scripting'. Look to the Qt docs and community for how to call OS APIs from your Qt app. (e.g. A common solution on Macs is to use QProcess execute an AppleScript via osascript; limited and slow, but sufficient for simple tasks.) 2. Use Adobe's CEP to embed a basic Node.js HTTP server in a custom panel which your app can then talk to same as it would a web service. Example here: https://medium.com/adobetech/how-to-build-a-node-js-server-in-a-panel-ba1d63ea67e2 ... View more