SIPRI (Stockholm International Peace Research Institute) has some databases available on weapons that were traded between countries or political groups. I think the weapons are going to be military-grade, so I'm not sure how useful for your question. Also, the data is there but quiet buried under manual extraction process and a format that is not easily machine readable.
Shows all international transfers in seven categories of major conventional arms since 1950, the most comprehensive publicly available source of information on international arms transfers.
The data at this level is aggregated, so for particular weapons, we have to go one level deeper.
Provide information on each deal included in the database. Information provided includes, inter alia, the suppliers and recipients, the type and number of weapon systems ordered and delivered, the years of deliveries, and the financial value of the deal.
The screen should look like this:
The exhausting effort will be to create all the combinations of countries, but you can narrow the search to countries that manufacture weapons as the supplier.
Then, when you've gone through all ∞ combinations, the data turns out to be in RTF files (which you can convert to text and then parse and structure).
Here's a screenshot of the "raw" data.
This is a good task for a web scraper like Python's mechanize, scrapy, etc