Beta package
pst-tool
Outlook .pst extraction done the way forensic and migration work needs it: every folder, every body, every attachment, out to plain files you can diff, hash and hand to somebody else.
Four small command-line scripts that run on your own machine. Nothing is uploaded, and the archive is opened read-only.
A beta package is unfinished software. Read the beta terms before you run it.
Download
In the browser, or on your machine
- In the browser — read an archive in the page. Choose a
.pstand it shows the folder tree, the message count and the first messages without uploading anything: the parser runs in the page against the bytes your browser reads off the file. On the test archive it reports the same 11 folders and 71 messages as the command-line tool. Compressed archives are the known gap, and the page says so when it meets one. - The sources — pst-tool.zip (15 KB): the four scripts, the README and the drag-and-drop batch wrapper, for when you want the whole archive written out rather than read.
- What is not in the zip: the dependency folder. The repository vendors it, but a 15 KB archive should not carry it, so the first step below is
npm install.
Verify what you downloaded
32dfade3261c3f30e64447ecc100d0c634a6ececeef180847c8a660f991431aashasum -a 256 pst-tool.zipRun it
One command in, a tree out
# 1. unpack and install the one dependency
Expand-Archive .\pst-tool.zip -DestinationPath .\pst-tool
cd pst-tool
npm install
# 2. the single command that does the job
node pst_index.js "D:\Mail\2015.pst" # -> .\pst-index-output\
node pst_index.js "D:\Mail\2015.pst" D:\index # or choose the destination
# 3. reconnaissance, if you want to look before you leap
node pst_probe.js "D:\Mail\2015.pst" # store name, folder tree, item census
node pst_dates.js "D:\Mail\2015.pst" # per-folder date range, year histogram
# 4. bulk extraction from a prepared list
node pst_extract.js my_list.json out\Exit codes are part of the interface: 0 an index was written, 2 bad usage, and non-zero when the archive cannot be read or when zero messages were extracted. A run that extracts nothing fails on purpose — an empty result that reports success is how a broken extraction ships.
Behaviour is tunable from the environment rather than from flags: PST_MAX_BODY caps the stored body length and PST_ATTACH controls attachment handling.
Status
What is finished, and what is not
- 26 tests pass, and they run against a real archive rather than a fixture stub.
- The read-only claim is tested, not asserted. The integration run hashes the archive before and after extraction and requires it to be byte-identical. A tool that quietly rewrites the evidence is worse than no tool.
- What a beta tester cannot do yet: the tool has not been run against a very large production archive on this side, so an unusually damaged or multi-gigabyte store is the case most likely to fail first. Tell us if it does.
Tell us what broke
Which archive shape failed, what the tool printed, and what the exit code was. Do not send the archive itself.