Various un(der)documented command line tools from my bin folder. Most should work on Mac OS X and Windows, and probably Linux too.

Most respond to -h.

Licence is GPL unless otherwise specified.

Convert a binary file into a list of C-style strings. You can paste the result into a Python or C program, copy it to a file and #include it into a C program, and so on.

Specify -, to have a comma after each line of input, giving you an array of strings, one per line; specify -n - probably most useful in conjunction with -, - to leave the newlines out of the strings.

List files to process on the command line, or specify no files to have it read stdin.

Convert a binary file into comma-separated hex values. You can paste the result into a Python or C program, copy it to a file and #include it into a C program, and so on.

Makes a pair of files, .h and .cpp/.c/.m/.mm, with some skeleton C++ code in. Automatically generates sensible header guard defines and, when generating C++ .h files, some markup so emacs knows they're not C.

There are various options, that you can supply on the command line.

To keep files consistent, mkhc will look in the current folder, and any parents, for a file, .mkhc. If it finds it, it will read additional command line options from it, one per line. The long form of each option should be specified, without the leading --. For example:

header-folder=h/shared
src-folder=c
c
extern-c

This is equivalent to running mkhc as follows:

mkhc --header-folder=h/shared --src-folder=c --c --extern-c

Lines beginning with # are comments.

By default, mkhc stops searching at the first .mkhc file it finds, but you can add the special magic comment #.. (just like that, on its own line, with no spaces) to have mkhc keep searching parent folders for more .mkhc files.

When specifying paths in a .mkhc, specify them relative to the folder the .mkhc is in.

(Run mkhc -v, possibly with additional options, to show the options that will be used, taking into account any .mkhc files read.)

Look after sets of branches named CATEGORY/NAME. List branches and categories. Find branches merged into another branch - candidates for deletion. Find branches not merged into another branch - candidates for force deletion, or possible recategorization.

Batch change of file extensions.

Prints input received on stdin, with thousands separators added to any numbers printed.

> p4 sync -N ...
Server network estimates: files added/updated/deleted=3913/1635/0, bytes added/updated=1593990200/1166508036
> p4 sync -N ... | commas
Server network estimates: files added/updated/deleted=3,913/1,635/0, bytes added/updated=1,593,990,200/1,166,508,036

Counts occurrences of contents of lines on stdin, and prints the tallies.

Move files into folders according to their creation date.

Does a folder diff without printing out too much verbiage.

Can optionally invoke diff on edited files, to get the verbiage back.

A version of the handy *DUMP command from the BBC Micro.

Look for apparent text file line endings (CR+LF, LF+CR, CR, LF) and print the number of each type found. If the lines aren't all consistent (seemingly the usual cause of git thinking the entire file has changed), use one of the appropriate command line options - see -h output - to print the line numbers of a given type.

Find duplicate files.

Reads a list of files on stdin, one per line, and executes a command for each file.

(This works like xargs, but it makes lighter work of files that have spaces in their names, and can work with the output from a wider range of tools, particularly on Windows.)

Makes up a random password that can be copied from iTerm2 with one double-click.

Modifies a file by changing every byte. For binary files, each byte is XORed with 255; for ASCII text files (use -a), it does a rot47 on the 94 printable chars. (-a may not actually change anything, of course, if the file isn't really ASCII.)

Splits an image into equally-sized tiles, possibly after removing border regions based on alpha channel. The tiles can optionally be generated overlapping, to avoid gaps due to transformation inaccuracies.

Dependencies: PIL

Sends stdin to emacs via emacsclient, so you can use emacs as the target of pipes.

If using recentf, add the following to your .emacs:

(add-to-list 'recentf-exclude "pmacs\\.[0-9]+\\.dat$")

Batch convert WAVs and FLACs to MP3s. Useful for devices that have limited storage, don't support FLAC, etc.

(lame is run with --preset insane, so 320Kbps output. This is compatible with all the playback devices I use.)

Dependencies: lame, flac tools (if converting FLACs), GNU make

Find MP3 "residue" (a poor choice of terminology, sorry!) - the difference between the signal in the original WAV and the signal in the MP3.

Specify the name of a .wav file. It will compress it to MP3 using various settings (both CBR and VBR), then for each set of settings it will produce a WAV that's the difference between that MP3 and the original WAV. The 128kbps one will probably be noisy; the 320kbps one will probably be very quiet.

Inspired by: http://ryanmaguiremusic.com/theghostinthemp3.html

Dependencies: GNU make, mpg123, lame, rm, flac (only needed if you use a flac file as input)

Makes a seamlessly-looping MP3 from a WAV file. Requires LAME (http://lame.sourceforge.net/).

For the principle, see http://www.compuphase.com/mp3/mp3loops.htm.

Lists WAV file chunks. Pretty-prints chunks it knows about.

Uses ffmpeg to create a video file from a sequence of image files and WAV files. Specific options for my favourite tasks: output 720p or 1080p, correct aspect ratio from b2 output, and crop to remove borders.

Mainly intended for creating videos from the output of b2, but could be multi purpose.

Dependencies: ffmpeg

I haven't done any iOS stuff for years. This stuff probably no longer works.

Takes asm output from Xcode on stdin, and prints it to stdout, only with .loc directives replaced with the actual lines from the original source code. This is what Visual C++ does, and it's very handy...

(To get asm output from Xcode, click the button at the top left of the text edit window - it looks like a little 4x2 grid, I've no idea what it's supposed to be - and select Assembly from the menu.)

You can copy the result from Xcode and use pbpaste to pipe it through improve-xcode-asm-output, e.g.:

pbpaste | ./improve-xcode-asm-output.py

Helper script for importing device IDs en masse into the developer portal without endless "this devices already exists" errors.

How to use:

1. Ask Test Flight to export device IDs list for your team members. You'll get a file called something like testflight_devices.txt in your downloads folder.

2. Visit the iOS Provisioning Portal, Devices section. Use File>Save As to save the page in Page Source format (i.e., an HTML file).

3. Run this script, supplying name of HTML file and devices txt. The output is all device identifiers and device names that are mentioned in the Test Flight devices list, but not in the provisioning portal. Redirect the output to a .txt file, since you'll need it in the next step:

   ./iosids.py /tmp/Devices\ -\ iOS\ Provisioning\ Portal\ -\ Apple\ Developer.html ~/Downloads/testflight_devices.txt > /tmp/devices.txt
   

4. Use the Upload Devices button in the Provisioning Portal to bulk add the new devices. Point it at the text file created in step 3.

Symbolicates a crash log from the iPhone. Apple supply a perl script to do this, but it relies on spotlight having indexed the dSYM folders for the binaries. It never seems to do that on my system, so the script always fails. I believe you can use Xcode to do it, too. But I don't like perl, and I don't like Xcode.

symcrash uses spotlight to search for dSYM files by file name, which appears to be perfectly reliable, then looks through all the dSYMs found to find the one for the binary in question.

(symcrash does not support as many different kinds of crash log types as Apple's perl script.)

Converts an Atari ST GEMDOS format executable file (.PRG/.TOS/.TTP) into a memory image suitable for use by a disassembler. It loads the program in, relocates it, puts the zero-filled BSS in the right place, then saves the result. Have your disassembler start from the first byte (which is the usual BRA *+$1E).

Extract files from Atari ST trackwise floppy disk images.

(May also be suitable for DOS disks, which have a very similar format, but that's untested.)

Split Atari ST ROM images into banks, for programming into multiple PROMs. Also verify ROM checksums present in TOS 2.06+.

Modify route tables when using VPN that takes over your PC's entire network interface. Removes VPN routes for internet and LAN, and (optionally) reinstates them for specified address ranges corresponding to whichever VPN resources you need to use.

Handle Visual Studio multithreaded build output.

Use list to list windows (optionally filtering by PID or window title) and their base64-encoded placement data.

Use set to set a single window's placement (filtering by PID or title is required) using the data printed out by list.

Intended workflow:

1. have batch file or whatever that runs a few programs that open windows in random places

2. run batch file and arrange windows in desired layout

3. use window_placement list interactively, to get the placement data for each window of interest

4. add window_placement set invocations to batch file (or wrapper of) so that the desired window layout is reproduced on each run

Does an svnsync from a remote repo.

Analyze Time Machine backups. There are various options, but just run it like this, from your Time Machine backups folder:

tma -wum

This tells you how the most recent backup differed from the penultimate one. Handy if Time Machine backs up a ton of stuff, and you're not sure why.

More than somewhat reminiscent of dtruss, but it's in Python, and the dtrace bits are in their own file so it can work a bit better with Emacs.

Also calls sudo for you.

Takes in output from dtruss or dtruss.tom, and outputs it annotated with the odd useful bit:

• for a few key syscalls that take an fd, note the path (when it can be determined)
• add symbolic names for whatever flags/modes/etc. I've found myself in need of

May not (yet...) parse every possible type of dtruss output.

Various text-processing scripts for use with a text editor.

For emacs, use M-x shell-command-on-region.

For Visual Studio 2012 and later, use my VSScripts addin from https://github.com/tom-seddon/VSScripts.

See each file for the licence.

See https://github.com/chromium/chromium/tree/master/tools/symsrc

https://github.com/erocarrera/pefile

Written with the intent of being useful, but, for now, at your own risk...

Invokes cdb to print a list of the modules in a Windows .dmp file.

Finds a Windows .exe or .pdb by embedded timestamp in a folder structure. This is the info used to find the .pdb that matches a .exe; when you have one of the pair, this might help you find the other.

For a bit more about this, see http://www.debuginfo.com/articles/debuginfomatch.html

Print some info about a .pdb file.

For an overview of the pdb format, see https://llvm.org/docs/PDB/MsfFile.html

Print some stuff - though not much, at least yet - from an EXE header.

This is a tiny little bit like dumpbin /headers EXE in a Visual Studio command prompt, but the output is a compact, one-line format, for easier interop with grep or findstr.

For an overview of the PE format, see https://en.wikipedia.org/wiki/Portable_Executable