`archive` pseudo-vcs driver: indexing code in archives (e.g. zip, tar) without extracting files

What kind of change does this PR introduce? (check at least one)

[ ] Bugfix
[x] Feature
[ ] Code style update
[ ] Refactor
[ ] Build-related changes
[ ] Other, please describe:

The PR fulfills these requirements:

[X] All tests are passing?
[X] New/updated tests are included?
[ ] If any static assets have been updated, has ui/bindata.go been regenerated?
[ ] Are there doc blocks for functions that I updated/created?

If adding a new feature, the PR's description includes:

[X] A convincing reason for adding this feature (to avoid wasting your time, it's best to open a suggestion issue first and wait for approval before working on it)

Description:

This PR adds a new driver archive, which allows to index source code in archives (e.g. zip, tar; any that supported by https://github.com/mholt/archiver) without extracting files: while indexing, files are walked using archive API, and while searching, results are checked and snippets generated with files extracted on the fly.

A config example:

{
  "dbpath" : "db",
  "vcs-config" : {
    "git": {
      "ref" : "main"
    }
  },
  "repos" : {
    "video" : {
      "url" : "/Volumes/1tb-ext4/twitch/video.zip",
      "vcs" : "archive",
      "vcs-config" : {
        "ignored-files" : [".git"]
      },
      "url-pattern" : {
        "base-url" : "file:///Volumes/1tb-ext4/src/twitch/{path}"
      }
    }
  }
}

Some metrics:

for 160 zip files, 126GB, I got 3GB of indexes
it takes about 13 seconds for a search request to execute

hound-search / hound

`archive` pseudo-vcs driver: indexing code in archives (e.g. zip, tar) without extracting files #484