Expected type is `"CortextDB"` and you should fill the tables to search in.
```json
"reldbs":{
"CortextDB":{
"file":"$$relpath/to/some.db",
"qtable":"$$tableNameToSearchIn",
"template":"cortext_with_link"
}
}
```
###### => for twitter queries
Expected type is `"twitter"` and no additional conf is needed (POSS for the future: add twitter query context, ex: "Présidentielles 2017 AND (query)").
```json
"reldbs":{
"twitter":{}
}
```
###### Real life examples
```json
"data/gargistex":{
"first":"shale_and_ice.gexf",
"graphs":{
"shale_and_ice.gexf":{
"node0":{
"name":"terms",
"reldbtype":"csv",
"reldbfile":"shale_and_ice.csv",
"reldbqcols":["title","abstract"]
"graphs":{
"model_calibration.gexf":{
"node0":{
"name":"terms",
"reldbs":{
"csv":{
"file":"model_calibration.csv",
"qcols":["title"],
"template":"bib_details"
},
"twitter":{}
}
}
}
}
},
"data/test":{
"first":"mini_for_csv.gexf",
"graphs":{
"mini_for_csv.gexf":{
"node0":{
"name":"terms",
"reldbs":{
"csv":{
"file":"mini_for_csv.csv",
"qcols":["title","keywords","text"],
"template":"bib_details"
},
"twitter":{}
}
},
"model_calibration.gexf":{
"node0":{
"name":"terms",
"reldbtype":"csv",
"reldbfile":"model_calibration.csv",
"reldbqcols":["title","abstract"]
"node1":{
"name":"authors",
"reldbs":{
"csv":{
"file":"mini_for_csv.csv",
"qcols":["author"],
"template":"bib_details"
}
}
}
}
}
}
```
###### => for CortextDB SQL tables
Expected type is `"CortextDB"` and you should fill the table to search in.
```json
"reldbtype":"CortextDB",
"reldbqtable":[]
```
In the last exemple, we have two nodetypes:
- node0 allows both CSV and twitter relatedDocs tabs.
This is a stub for a future documentation for developers.
#### About settings
- system-wide settings are in `settings_explorerjs.js`
- source-by-source settings (nodetypes, relatedDocs APIs) are in `db.json`
## Graph input choices
...
...
@@ -13,7 +16,7 @@ Tina allows 3 main ways of input :
The `sourcemode` value is by default the one in settings_explorerjs.js (`TW.conf.sourcemode`), unless an url argument of the same name is present.
The `serverfile` option has an extended version called `servermenu`. It opens a list of files called`db.json` on the server, providing a menu to choose from it.
The `serverfile` option has an extended version called `servermenu`. It opens the list of files from`db.json` on the server, providing a menu to choose from it.
The detailed implementation of these choices can be found in the function `syncRemoteGraphData()` in main.js.
...
...
@@ -32,7 +35,8 @@ This will still evolve but the main steps for any graph initialization messily u
4.[`main.js`] mainStartGraph() function runs all the rest
2. calls [`sigmaUtils`] where the function `FillGraph()` was a central point for filtering and preparing properties but now with 2 and 3 it just creates a filtered copy of the nodes and edges of the current active types to a new structure that groups them together (POSSIBLE remove this extra step)
3. back in [`main.js`], finally all sigma settings (user + defaults) are merged and we initialize the sigma instance (`new sigma` etc.)
3. back in [`main.js`], finally all sigma settings (user + defaults) are merged and we initialize the sigma instance (`new sigma` etc.)
at this point, any additional conf located in db.json is used for nodeTypes and relatedDocsTypes
4. finally a call to [`TinawebJS`] initializes the action listeners and this phase should crucially initialize items that need the sigma instance (because they may depend the displayed categories, the number of displayed nodes, etc)
