$.line and $.stack counted lines in the run's expanded source. In code
that comes after an #include they gave lines that match no file, and
this included code in a file $.evalFile ran and code a named engine
kept from an earlier run.
Both now map each frame through the run's source maps, as uncaught
errors already were. $.stack writes each frame as name@file:line,
outermost first, with the run's name as the file of a script given as
source. Code that eval or Function compiled keeps the line of the
source they compiled, in [eval]; code whose place a named engine no
longer keeps is at line 0 of [unknown]. The 64-frame limit, [global]
and [anonymous] are unchanged.
Docs: docs/mcp.md, docs/target-app-parity.md.
flat_charged and the property-key path now root the rope itself, so a string held only by
native code (a getter's result for ++o.x, a valueOf result, a toString key) still counts
what it holds when the reservation collects. The per-call roots for these cases are removed.
When the flattened copy of a String object's concatenated string would pass the heap limit
before a collection can run, the indexed read walks the rope to the unit instead, taking a
unit of fuel for each node it visits, so no copy is made.
- wrapping a string no longer flattens it, so Object(s).length and new String(s) on a large concatenated string make no copy and cannot collect while a call's operands are off the stack
- reading a String object's character by index, `in` and hasOwnProperty count the flattened copy at the next limit check, and the run ends before the copy when the fuel for it is not left
- unary plus on an object whose valueOf returns a concatenated string, and a property key from toString, keep that string rooted while its copy is reserved
- the unshift comment cites step 6 of section 15.4.4.13
Boxing the string receiver of a non-strict function flattens it, and reserving the copy can
collect. The callee, receiver and arguments are already off the stack at that point, so they
stay in the temporary roots until the frame holds them; before, an argument object could be
freed and its cell reused by the wrapper. new String(value) roots the converted string the
same way, so what the rope holds is still counted while its copy is reserved.
- the relational operators root both strings while flattening them, so the first copy still counts when the second one starts a collection
- join roots its separator for the whole join and each element while it is written, and JSON.stringify roots a string while it writes it
- a String object flattens its string when it is made, with the copy counted, so reading one of its characters copies nothing
A for each loop read the property's value before evaluating its target,
so a target such as (o.p = 2, t).v, t[(o.p = 3, 'w')] or a bare
@[(o.p = 2, 'k')] stored the value from before the target ran. The
target's object and name are now evaluated first and the value is read
after them (ECMA-357 section 12.3).
Flattening a concatenated string copies its units into a new buffer. The
string built-ins reserve that copy with flat_charged, but several other
paths flattened without it, so a script could build copies the heap limit
never saw:
- the relational operators (`<`, `>`, `<=`, `>=`) on two strings, `==`
between a string and a number, unary `+` and every other ToNumber of a
string, and reading a character by index (`s[0]`)
- a property name made from an object whose toString returns a long
string, and a long string used as a property name, which now reserves
its copy before it is made instead of after
- direct `eval` of a concatenated source
- `JSON.stringify` of a string value and its `space` argument
- `Array.prototype.join`'s separator and each string element it writes
Property names and eval source reserve the copy's heap bytes without
taking fuel.
Direct and indirect eval, Function and host evaluators parsed and compiled
their source before taking any fuel or checking the heap, so a script with a
small fuel limit could make the interpreter parse megabytes of source in a
handful of instructions. Interp::prepare_source now takes one unit of fuel for
each UTF-16 code unit and checks that the source text fits in the heap before
the source is flattened, parsed or compiled. A host evaluator's string takes
the fuel for a third of its bytes before it is encoded, and the rest after.
Limits::fuel and the maxFuel row in docs/mcp.md say so.
With Options::e4x set, an E4X name standing alone (@a, @ns::a, @[expr],
ns::a, *) is now a valid target for =, compound assignment, ++, --,
delete and for/for each loops, as well as a read. The target is the
nearest XML or XMLList value in scope whose [[HasProperty]] accepts the
name (ECMA-357 section 11.1). The name is evaluated once, and the
target resolves before the right-hand side runs.
When no XML value in scope holds the name, a read or write throws a
ReferenceError and creates no child or attribute. A read no longer
falls back to the first XML value in scope, so a filter such as
list.(@id == '1') throws when an item has no @id and no outer XML
value holds one, instead of treating that item as not matching.
A finally block that break or continue leaves no longer keeps the saved
site of the exception it ran for. A Jump now removes the sites at or
above the stack's length, as a return already does, so a collection
after the stack grows back over that slot does not root and count the
old code.
New tests in crates/js/src/host_tests.rs: an error passed on by a
finally block that collects garbage under a 1 MiB heap is reported at
its throw; the saved site keeps direct eval code live through a
collection; a host abort inside a finally block leaves no saved site;
and break, continue and return out of a finally block drop the site.
The saved throw sites of running finally blocks are kept in rising slot
order, so leaving a frame or a block now removes entries from the end of
the list and stops at the first one below the slot. Before, every return
looked at every saved site, so a call made inside many nested finally
blocks took time in proportion to their count.
A `return` inside a finally block discards the pending exception, but the
saved throw site for it stayed in the list of collection roots. The code it
pointed to then stayed live after the function was gone, which could stop a
script with the heap limit when that code came from `Function` or `eval`.
Returning from a frame now drops the saved sites for the slots it leaves.
Follows #6653.
random_seed passes the OS random source's result to random_seed_from, so a test can
give it a known value and check the seed it makes, instead of only reading the source text.
An error a finally block passed on unchanged was reported at the end of
that block, in script.run's report and in the Script Console, because the
block's end raised the error again at a new site. The interpreter now keeps
the site of the exception each finally block runs for, keyed by the stack
slot that holds it, and the block's end throws the exception on from that
site. This holds through nested finally blocks, called functions, resumed
runs and marked host errors. An error the finally block throws is reported
at that throw, and one its return, break or continue discards is not
reported.
Each realm seeded Math.random from std's RandomState. On
wasm32-unknown-unknown its keys come from stack and heap addresses,
which are the same on every page load, so scripts in the web build
drew the same sequence each time the page opened.
The seed now comes from getrandom (crypto.getRandomValues in the
browser, as crates/pdf/src/encrypt.rs already does), with RandomState
kept as the fallback when it fails.
Scripts get the File and Folder classes as built-ins of the script
engine. A script run from a file can test whether files exist, read
their properties, list folders with getFiles (a mask or a function),
and open files to read them with read, readln, readch, seek and tell.
- Reads stay inside the script's folder and, when automation read
roots are set, inside them too. A script given as text reads no
files. Network and device paths, and paths outside the folder by
their names, are refused before anything is looked up. Links and
junctions that lead out are refused, and getFiles leaves them out.
- The encoding comes from the byte order mark, then the encoding the
script set, then UTF-8 when the bytes are valid, then CP1252.
- Limits: 8 MiB a file, 32 MiB a run, 256 files open at once, 2,048
getFiles entries and 65,536 names looked at, and paths of at most
32,767 characters.
- Writing, the file dialogs and execute are not available yet: they
return false or null and set error.
The report's hostCallCounts gains fileCalls and fileExecutes.
Docs: docs/mcp.md, docs/target-app-parity.md.
An exception that a finally block passes on, or one that leaves a
built-in's callback, no longer becomes $.error again over an error thrown
after it; throwing a caught error again still makes it the last one.
$.sleep with more than 0 inside a getter, a conversion or a built-in's
callback now throws whether or not the platform has a clock, so the web
build throws there as the other builds do.
A marked host error whose finally block was left by return, break or
continue kept its entry, and its error object, until the run ended. A
script that discarded many of them reached the heap limit. Each entry
now records the stack slot that holds its exception, and entries whose
slot no longer holds it are removed before the next exception is handled
and before each collection.
A failure $.evalFile raised for a file it could not run was placed by
matching the uncaught error's message against the last such failure. A
script that caught the failure and threw it again later, or threw a new
error whose message held its text, was reported at the failed file's line
instead of the second throw.
The failure now carries a mark (HostError::Marked) with the exception.
A catch clause drops the mark, a throw raises the value without one, and
a finally block that ends normally throws it on with its mark, nested
finally blocks included. Realm::error_mark gives the mark of the error a
run ended with, and the engine looks up the file, line and column by it.
`#strict on` and `//@strict on` no longer end the run as unsupported. They
turn on the checks the dollar-object help page gives for `$.strict`
(https://extendscript.docsforadobe.dev/extendscript-tools-features/dollar-object/):
an assignment to a read-only property throws a TypeError, and so does a new
property on an object that takes none.
VectorCraft's own rules: a getter-only property counts as read-only; the
objects that take no new properties are the non-extensible ones
(preventExtensions, seal, freeze, and the frozen vc and $ objects); the last
`#strict` line of the script and its includes sets the whole run, and runs
without one are unchanged; eval and Function take the checks of the code
that compiles them; a file $.evalFile runs keeps its own line's checks for
its code and functions, or takes the caller's without one. The rest of ES5
strict mode stays with 'use strict'.
The js crate carries the setting per compiled source
(Realm::set_strict_writes, and a strict_writes callback on
HostItem::Evaluator), and applies it only to property assignments.
A script's $ object now has the other members the Dollar object page
lists: about, appEncoding, bp, build, buildDate, colorPicker,
decimalPoint, engineName, error, flags, gc, getenv, setenv, global,
hiresTimer, level, line, locale, localize, memCache, os, screens, sleep,
stack, strict and version. Scripts also get the global
localize(object, ...values), toSource(value) and
Object.prototype.toSource().
Each run starts with the interface's locale and its own $ state;
$.setenv changes only the run's copy of LANG, LC_MESSAGES, LC_ALL and
OS, and $.getenv returns null for every other name. $.sleep waits from
0 to 1,000 ms between frames, and script.stop ends the wait. toSource
writes plain data only and throws on cycles, getters, host objects and
values past the data limits.
The script checker reports $.localize as a $ member instead of the
global localize function.
alert(message[, title, errorIcon]), confirm(message[, noAsDflt, title])
and prompt(message, preset[, title]) are globals in a script run. In the
app the run waits at the call for the person's answer, spends no
instructions while it waits, and every dialog has Stop, which ends the
run and puts every document back. A run over MCP or the command line
takes its answers in order from script.run's answers, a list of
{kind, value} records; a missing, mismatched or unused answer ends the
run. The report lists each dialog with its answer in dialogs, with
dialogCount and dialogsTruncated.
The interpreter gains a host call that waits at the call for its
result, from the script's own code only; a dialog called in a callback
of a built-in, a getter or a conversion throws in the app.
Docs: docs/mcp.md, docs/target-app-parity.md.
Before a script runs, `script.run` reads its preprocessor lines (`#` or
`//@` at the start of a line, outside strings and comments):
- `#include` and `//@include` put another file's text in place of the
line, looked for beside the including file, then in each
`#includepath` folder, then in each `JSINCLUDE` folder. A later
`#includepath` replaces the list.
- `#script` names the run unless `name` is given; `#target` takes any
value; `#strict off` is taken. `#strict on`, `#targetengine` and
`#engine` are refused as unsupported, and any other `#name` as
unknown.
`$.evalFile(path, timeout?)` runs a file in the same global environment
and returns its last value, `$.fileName` is the file being run and
`$.includePath` the include folders.
A script run from a file reads these files only in its folder and the
folders below it. Paths with a `..` name, links or junctions that lead
out, device names, streams and network paths are refused; a script
given as text reads none. A run reads at most 64 files 16 levels deep,
32 MiB of source in all and 1,024 paths looked up. An error in an
included or evaluated file is reported at that file's line and column,
and the report's new `file` field names it.
Docs: docs/mcp.md (script.run, Script Console), docs/target-app-parity.md.
Scripts write to a new Script Console with $.write(...values) and
$.writeln(...values), each value as String(value) gives it. vc.log
entries, a run's uncaught error as name:line:column: message, and a
start, end and report line for each run go there too. The report line
gives the run's time, host calls and slowest commands.
- The console keeps at most 2,048 lines and 1 MiB of text, dropping the
oldest and counting what it dropped. A line is cut at 4,096 bytes and
a script's name at 256 bytes, with a count of cut lines. Line feeds,
carriage returns and CRLF break lines; other control characters show
as Unicode Control Pictures.
- script.run's result and scriptReport gain runTimeMs, timingAvailable,
hostCallCounts, slowestCommands and the run's own console. A failed
run's report has them too.
- script.console and script.console.clear read and empty the console;
ui.copyScriptConsole {first?, last?} copies lines as the panel shows
them.
- The panel lays out only the rows in view, selects with click and
Shift-click, and has Clear, Copy Selected and Copy All. Output goes on
while it is closed, and Clear works during a run.
Docs: docs/mcp.md, docs/gaps.md, docs/roadmap.md,
docs/target-app-parity.md.
File › Scripts › Other Script… picks a .js or .jsx file, shows a warning
that can be turned off in Preferences › General, and runs the script
against the open documents through the `vc` object. The new
`script.run` and `script.stop` commands run and stop a script over MCP.
- A run is one transaction. Each document it changed gets one undo step
named Run Script. An uncaught error, a limit or a stop puts every
document, the paint, the clipboard, menu state and preferences back.
- A command that fails inside a run, or whose result is too large to
reach the script, is undone and throws an error the script can catch.
- Files a command wrote stay on disk and the report names them, also
when the run is stopped from the quit question.
- While a script runs, other commands that change anything fail with a
busy error; MCP reports it as JSON-RPC error -32003.
- In the app the run advances one slice per frame with a Stop button in
the status bar. A run that adds a great many objects asks whether to
continue; Stop and Escape end it.
- Commands that change state outside the documents, such as prefs.*,
plugin.* and file.place.queue, are refused to scripts.
Docs: docs/mcp.md, docs/gaps.md, docs/roadmap.md,
docs/target-app-parity.md.
With `Options::e4x` set, vectorcraft-js runs ECMA-357 (2nd edition):
XML and XMLList literals with embedded expressions, the `..`, `.@`,
`.*`, `.()` and `::` selectors, assignment and `delete` on selected
nodes, `for each`, `default xml namespace`, and the four objects with
their methods and settings. The option is off by default, and without
it the syntax and globals do not exist.
Parsing, serializing, comparing and copying trees use no native stack,
and each takes fuel and counts against the heap limit. Inserting a node
into itself or its descendants throws a TypeError before the tree
changes. A literal or XML string nested past `Options::max_depth` gives
a SyntaxError, and a tree built past that depth stops at the heap limit.
`XML.prototype.xpath` is not provided.
Installing the built-ins is not bounded by the heap limit: Math and JSON
no longer reserve heap while they install, so a realm made with a tiny
heap_bytes gets every built-in and runs once the limit is raised. If
installing fails, the realm keeps the reason and every run ends with an
internal error instead of running against missing built-ins.
The json2.js test skips only when the built-in check returns false, and
fails on any other result. The parser's progress line says the app does
not run scripts yet.
A realm made with a low fuel or heap limit lost RegExp and Date, because
installing them compiled a pattern under that limit and the error was
dropped. The built-ins now install before the realm's limits apply.
The early pattern check read an octal class escape starting with 4 to 7
as three digits, while the compiler reads two, so `/[\477-z]/` gave a
SyntaxError for a pattern the compiler accepts. The check now reads
class octal escapes with the compiler's own reader.
docs/target-app-parity.md: the interpreter's entry lists the new
built-ins and the current language-test count.
dotCooCoo