Activity for John-Philip Taylor
-
John-Philip Taylor posted a comment on discussion General Discussion
I've been threatening for ages to implement a "use a script" feature, so that you don't spec everything on the command-line, but rather a YAML file or something. Something like: gerber2pdf -script myscript.yml That way you eliminate the command-line limitation. Maybe the time has arrived...
-
What is the error? My suspicion is that it is a command-line command length limit of whatever shell you're using. I "think" Windows CMD has a limit of 8191 characters or something, but it varies depending on exactly what version of Windows and / or shell (CMD vs PowerShell, for example) you're using. Browsing through the source code: The number of commands on the command-line (limited by Gerber2PDF, not the shell) is 4096 commands. You should be able to pass in more than a 1000 files without issues,...
-
New version is ready for you to download... The command-line option you want is the "-name_is_filename" option
-
Added the "-name_is_filename" command-line option
-
Changed indentation from 2 to 4 spaces
-
Changed indentation from 2 to 4 spaces
-
I'm sure that I can add a command-line switch for that one quite easily... I'll see what I can come up with over the weekend...
-
Fixed the Pascal to Atmosphere conversion
-
Bugfix aperture indices >999
-
Updated the submodules and switched to C++20
-
I fixed the larger aperture indices (I've changed the array to a map, so it can now handle aperture indices anywhere in the "int" type range, i.e. 2 x 10^9). I'll upload the new version shortly. I cannot reproduce your second problem, however... It produces the attached PDF.
-
No problem -- I'll have a look over the weekend...
-
Sounds like an easy fix... I'll have a look before the end of the week...
-
More EBNF, parser and AST updates
-
Updated the readme
-
More dangling comma support
-
Combined TypeIdentifier and CoverBinIdentifier and added dangling comma support
-
Improved HDL construct handling in parser and AST
-
Added fence nodes after assignments and function calls
-
Merge branch 'interpreter' into develop
-
The readme now points to the local folder instead of SourceForge
-
Updated the Wiki to bring it up to date with the latest design and ideas
-
Added coverbins and covergroup constructs to the parser and AST
-
Fixed consistency of parameter and attribute list ordering
-
Ripped out labels and goto's
-
Added enum attributes and initialisers to the parser and AST
-
All AST parameters in the interpreter are now const
-
Improved import path search algorithm
-
Added a clarification on imports to the wiki
-
Basic import is working
-
Code clean-up:
-
Major update to the Number class
-
Renamed the main executable to all lower-case
-
Added assertion sequences to the parser and AST
-
Moved the Wiki readme to the Wiki root folder
-
import can now take a stringification expression as a filename
-
Cleanup of the test script globals
-
Added variable initialisation to the interpreter
-
Improved the interpreter architecture
-
Merge branch 'newAST' into develop
-
Moved TokenTree to FrontEnd
-
Updated the SVG figures to work on dark-mode readers
-
Added the Alcha case study code as a parser test
-
Added parser tests and bug-fixes
-
Added pragmas to ignore the format-truncation warnings
-
The parser and AST now support interpolated strings
-
EBNF update: interpolated strings
-
The scanner now supports interpolated strings
-
Implemented the rest of the AST (untested)
-
Set up the new AST classes for statements
-
Implemented all the `Verification` EBNF
-
Implemneted all in the `Declarations` EBNF
-
Expression parsing implemented
-
Split the Parser and Interpreter tests
-
Moved the AST::run() functions into a central Interpreter class
-
"num" declarations are working (minimal implementation)
-
Basic print now works via an AST instead of directly
-
Need to bring back the AST
-
EBNF: Cleaned up FunctionDef, OperatorOverload and IdentifierList
-
EBNF: The `inline` keyword now goes in front of the return type
-
Added `test%` targets to the tests Makefile
-
Ripped out the AST, Netlist, Engine and BackEnd, and `print` is working
-
Merge branch 'code_cleanup' into develop
-
Added an automated Scanner test
-
Scanner::getToken now exits early on error
-
Refactoring:
-
More consistent naming convention in the enumerations
-
Updated build instructions
-
Upgraded GMP
-
Fixed so that it compiles on Linux
-
Coding standard update:
-
Updated the Toolbox to the latest version
-
Coding style update
-
EBNF Refactor: Case to statement and Postfix split
-
EBNF update: HDL pins can now have direction
-
EBNF update: Attribute lists are now primary expressions
-
HDL files are now optional
-
Added operator overloading to the EBNF
-
Corrections to the EBNF
-
Pulled in the latest list of HTML character names
-
Upgraded to compile with C++20, 64-bit
-
Done -- You can find the new release on: The download page Github
-
The aperture macro rotation argument is now optional
-
The problem is that the converter doesn't support optional macro arguments. This is a relatively easy fix, so I should be able to post an update shortly... But first: some coffee ;-)
-
Sure - I'll have a look over the weekend...
-
Awesome news! You need to tell the other guys to update their settings ;-) And thanks for the compliment.
-
Refactored the bug report string into a global constant and added the GitHub issues link
-
The problem is that your drill file has no header, and as such does not define any tools or coordinate formats or anything. For your drill file specifically, it defines the tool sizes as you use them, as part of the tool selection instruction, but even that is without any way of knowing what the units are. For Gerber, you have to define all the "apertures" before the first draw instruction, so it will be quite a significant re-write to detect this scenario, then read through the whole drill file,...
-
Thanks for the report - I'll have a look and get back to you...
-
Merge branch 'feature/IndividualPageSizes'
-
Added next_* flavours of page_size, orientation and scale_to_fit options
-
Merge branch 'bugfix/JonThe640/TearingPlanes'
-
Arc endpoints are now forced (in the PDF) to be what the Gerber specified
-
Increased the error tolerance for single-quadrant arcs
-
Changed the Makefile to explicitly link the static ZLib
-
Merge branch 'feature/zlib'
-
Updated the Readme with the change log
-
Changed JDeflate to use the ZLib library
-
Added ZLib as a submodule
-
So... Any luck in getting a Gerber file to me that causes the issue so that I can investigate the problem?
1 >
