Repository File Structure
Qlik Cloudsupported Qlik Sense Enterprisesupported Git serverrequired
A repository created by Gitoqlok looks like this after a commit that includes the load script, app objects and attached files:
README.md
loadScript.qvs
data_load_script/
Main.qvs
Auto-generated section.qvs
tab_order.txt
objects/
properties.json
appprops.json
fields.json
glossarylinks.json
sheets/JGVJR.json
measures/XxRUjAb.json
dimensions/03c94547-….json
masterobjects/PEtFR.json
variables/….json
bookmarks/….json
stories/….json
snapshots/….json
automations.json (Qlik Cloud, when committed)
dataconnections.json (when committed)
appcontent/
ABCLogo.png
reload_tasks/ (Qlik Sense Enterprise, when committed)
build/
<appId>.qvf (when the QVF export is enabled)| Path | Content |
|---|---|
README.md | Generated description with the app name, the linked app ID and a table Branch name → Qlik application (unless disabled in Repository Settings). The repository description on the provider also holds the app ID — do not edit it |
loadScript.qvs | The complete load script, all sections joined. Optional (Do not commit full loadScript.qvs file) |
data_load_script/*.qvs | One file per script section, named after the section |
data_load_script/tab_order.txt | The order of the sections. Optional |
objects/properties.json | App properties (title, description, theme, …) |
objects/appprops.json, fields.json | Additional app-level objects and field metadata (tags, comments) |
objects/glossarylinks.json | Links between master items and business glossary terms (Qlik Cloud) |
objects/sheets/<id>.json | Sheet properties and the full property tree of every chart on it |
objects/measures, dimensions, masterobjects | Master items, one file per item |
objects/variables, bookmarks, stories, snapshots | One file per object |
objects/automations.json | Qlik Application Automations selected at commit time |
appcontent/ | Files attached to the app: thumbnails, images, other uploads |
reload_tasks/ | Reload task definitions (Enterprise) |
build/<appId>.qvf | Binary export of the app |
File names are the Qlik object IDs, so a file keeps its name across commits and renames of the object show up as content changes, not as new files. JSON is pretty-printed by default; above the size threshold in Repository Settings Gitoqlok proposes minified JSON.
Reading the history outside Gitoqlok
Because everything is plain JSON and QVS, the standard tools of the Git provider work: blame, file history, pull-request diffs, search. A reviewer can see which measure expression changed in which commit without opening Qlik.
