Smart Cache

Smart Cache is a fast pure Java HTTP proxy server with offline browsing support.

It has URL blocking, URL rewriting. Cached files has not headers inside and cache directory structure copies closely remote server. Supports also large caches.

Features:

Transports - Support for HTTP/1.1 and HTTP/1.0 protocols
- Support for HTTPS protocol (CONNECT method)
- Support for clients talking HTTP/0.9 protocol
- Support for persistent connections in HTTP/1.0 and HTTP/1.1 protocols
- Client and Server side of persistent connections are independant
- Support spaces in URL
- Support for chaining with password protected homepage and ftp-proxy
Quick abort - Can continue downloading after client closes connection
- configurable by min. bytes, max. bytes, percent ratio
- configurable by max. time needed for finishing download on aborted connection
Security - Bind address configurable
- Allowed IP list (with netmasks) configurable
- Maximum of connections configurable
- You can restrict target ports for CONNECT method
- You can configure maximum request (and reply) size
Portability - Needs only pure Java 1.1 or higher
- Do not uses unicode conversion of jdk1.1 where possible
- Auto runtime check for different filesystem capabilities
- Log files uses line termination of target platform
- MacOS
- OS/2
- Windows 95,98
- Windows NT,2000,XP
- Linux
- Novell Netware
- FreeBSD
- Solaris
- SGI Irix
Offline use - Auto detection of disconnected use
- Support for fetching non cached pages via Smart Cache loader
- Support for refreshing cached pages via Smart Cache loader
- Normally non cacheable pages can be still browsed when offline
- When offline, returns a cached copy of page (when available)
- You can force offline mode by -local switch
- You can configure special hook for download manager in offline mode
Privacy - Referer header faking/replacing
- User-agent header faking
- Cookie filtering
- URL filtering
Data storage - Smart Cache do not use auto generated hashed filenames
- NO http headers or other junk inside stored files. There are ready for use.
- Directory structure of cached files copies structure of remote server
- Top level directories are hashed to 2 levels deep matrix
- Files are saved under their original names where possible
- Files keeps at least their original extension when they can not keep original name.
- Smart Cache can compress stored files which saves you about 50% of diskspace
- Meta information of cached data can be rebuilded without data loss
- Meta informations are stored on per-directory basis
URL filter - Two regexp styles: simple dos-like and complete
- Blocked content can be replaced with image of user-defined color, transparent image, empty page or with any URL
- Auto detection of changes and reloading of blocklist
- Support for black/white lists
- Blocked Javascripts can be replaced with empty file (not by image)
Referer filter - Can drop Referer header
- Can replace Referer header with fixed string
- Can replace Referer header with requested URL
Redirector engine - Two modes of operation: internal redirect / external redirect (sent back to the browser)
- Auto reloadable configuration files for both types
- Advanced rewriting possible via multiple wildcard rules
- Fast redirects can be implemented via rules
Cookie filter - Two modes of operation: incoming or outgoing
- Auto detection of changes and reloading of blocklist
- Can turn persistent cookies into non persistent
Quotes - Entire Sri Isopanishad included
- Selected Bhagavad-Gita verses included
- Random quote feature
- Prev. / Next. quote
User interface - runtime change of console loglevel
- runtime change of internal server tracing
- cacheable searches for directory in cache
- browse your cache content
- on-line deleting files from cache
Garbage collection - Never runs auto-magically
- Configurable with MANY options
- LRU based
- LRU value can be fine-tuned based on object type or URL mask
- low/high mark diskspace support
- configurable loglevel
Download manager hook - No special requirements for download manager. Any program which can read queue from plain text file can be used.
- Can hook into specific URL masks.
- URL Masks can have several types: online masks, offline masks, ignore masks.
- Can hook into specific Content-Type.
- Content-Type masks can have minimum filesize specified (do not download short files) and Content-type-URL-ignore mask.