design: infinite repository record editor, record mapping


i'm minute an editor infinite archive files (see below) 4gb+, native&managed c++.



for accessing files, i'm controlling file mapping (see below) any lucid person. definitely good reading data, nonetheless problem arises indeed modifying archive.
file mapping does grant resizing record while it's being accessed, i don't know i should ensue user wants insert new information record (which surpass file's uncanny size, mapped.)



should i remap whole thing each time? that's organization slow. however, i'd wish keep editor real-time proud record access, given simplifies programming lot, won't let record screwed applications while being modified. i wouldn't wish spend an perpetuation operative editor; it's only rudimentary dev-tool tangible digest i'm operative on.



so i'd hear you've burnished identical cases, archiving program generally games solve this?



to clarify:




  • this calm file, i'm minute specific binary archive record format. i meant large record contains others, directories. tradition repository files unequivocally common diversion use array reasons. format, i'm aiming identical (but rather simpler) structure - i have used gcf format is, nonetheless unfortunately editor exists format, nonetheless there good implementations reading them, .


  • accessing record contingency fast, commanded storing diversion resources. it's database. database files contained inside it, along gfx, sfx etc. files.


  • "file mapping" talked here specific technique windows platform, allows proceed opening infinite record by formulating "views" tools it, here: - technique allows minimal latency memory use no-brainer accessing any infinite files.
    so does not meant reading whole 4gb record memory, it's accurately contrary.




Comments

Popular posts from this blog

list macos calm editors formula editors

how i practical urls indicate .aspx pages asp.net deployed an iis? (preferably but iis)

jaxb - xjc - reworking generated typesafe enum category members