0

Application: I wish to publish a web-application that takes input strings, searches for the string in about 5,000 plain-text files and returns the name of the files with matches. Each text file is about 4MB (uncompressed).

Problem: In PHP, I could use exec(grep -l pattern dir/* ) and get the job done. However, for cost reasons, I would go for a shared web-hosting plan which normally do not allow for executing programs.

Could you please suggest any alternative to grep for web environment?

I have understood following so far:

  1. A binary program file for any grep-alternative (e.g sift) could work. However, the problem of executing on a shared server would remain.

  2. PHP function preg_match is inappropriate considering a large number of files and their size.

I am open to implementations of grep-like function in other languages (e.g perl or javascript). However, I am not sure if the performance would be comparable to grep and whether the problem of execution would still remain.

I have tried looking for different web-hosting providers and understood that a virtual-private server (VPS) might be the solution. However, the price for a VPS plan by all hosting providers I have come across is unaffordable.

Any solutions or guidance for this problem?

Rahim Rajwani
  • 23
  • 2

1 Answers1

0

Possible solutions depend on what your hosting provider offers and your budget. Will you have a RDBMS available? You could then use full texts search which many offers. If not you could use SQLite, which has full texts search support.

If you have to stick to low tech solutions, than the PHP solution linked on the right might work for you.

Perl has a File::Find module, which you could use.

jira
  • 3,890
  • 3
  • 22
  • 32