You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Commit d384a67
Browse filesBrowse the repository at this point in the historyBrowse files
Copy file name to clipboardExpand all lines: docs/manual/configuration/basics.md
+26-21Lines changed: 26 additions & 21 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -47,16 +47,16 @@ Listeners define how an adapter receives input. They are placed inside a `<Recei
47
47
<ApiListener
48
48
name="inputListener"
49
49
uriPattern="booking"
50
-
allowAllParams="false"
50
+
allowedParameters="comma,separated,listOfParams"
51
51
method="POST"/>
52
52
</Receiver>
53
53
```
54
54
55
55
Key attributes:
56
56
-`uriPattern` — URL path segment (accessible at `/api/{uriPattern}`)
57
57
-`method` — HTTP method (GET, POST, PUT, DELETE)
58
-
-`allowAllParams` — whether to accept all query parameters
59
-
58
+
-`allowedParameters` — A comma separated list with query parameters that may be used
59
+
-`allowAllParams` — Disable the use of query parameters completely
60
60
## Pipeline and Exits
61
61
62
62
A pipeline contains an `<Exits>` block and one or more pipes. The `firstPipe` attribute specifies the entry point. Without `firstPipe`, pipes execute sequentially.
@@ -75,7 +75,7 @@ Each exit has a `name`, a `state` (SUCCESS or ERROR), and an HTTP response `code
75
75
76
76
## Pipes
77
77
78
-
Pipes are the processing units inside a pipeline. Each pipe has a `name` attribute and one or more `<Forward>` elements controlling flow. For a complete and up-to-date list of all available pipes and their attributes, see the [FF! Reference](https://reference.frankframework.org/#/components).
78
+
Pipes are the processing units inside a pipeline and have control over the processing flow. As such each pipe must have an unique name using the `name` attribute and one or more `<Forward>` elements. For a complete and up-to-date list of all available pipes, possible sub-elements and their attributes, see the [FF! Reference](https://reference.frankframework.org/#/components).
79
79
80
80
### EchoPipe
81
81
@@ -109,7 +109,7 @@ Validates the incoming XML message against an XSD schema. See the [XmlValidatorP
109
109
110
110
### SenderPipe
111
111
112
-
Wraps a sender to communicate with external systems (databases, services, etc.). See the [SenderPipe reference](https://reference.frankframework.org/#/Components/Pipes/SenderPipe) for all available attributes.
112
+
Wraps a sender to communicate with external systems (databases, services, etc.). Unlike a `Pipe` senders do not control the flow of execution and may not be placed directly inside a `Pipeline` but must be wrapped by either a `SenderPipe` any type of `IteratingPipe`. See the [SenderPipe reference](https://reference.frankframework.org/#/Components/Pipes/SenderPipe) for all available attributes.
113
113
114
114
```xml
115
115
<SenderPipename="insertBooking">
@@ -122,7 +122,7 @@ Wraps a sender to communicate with external systems (databases, services, etc.).
- The query-parameters in the SQL query may reference a specific parameter using `?{name}`
224
+
- You do not need to specify the `datasourceName` attribute if you wish to use the default.
225
+
225
226
## Database Initialization with Liquibase
226
227
227
228
Frank!Framework uses Liquibase for database schema management. Enable it by setting `jdbc.migrator.active=true`.
229
+
Each configuration may use their own changelog file. It is possible to change the name of the file and also to change the datasource it needs to apply the changelog to.
228
230
229
231
### DatabaseChangelog.xml
230
232
@@ -259,8 +261,8 @@ Place this file in the configuration directory. It defines schema changes as ord
259
261
260
262
Important rules:
261
263
- Never modify existing changesets after they have been applied. Always add new changesets.
262
-
- The default datasource when using Frank!Runner is `jdbc/${instance.name.lc}` (e.g., `jdbc/frank2manual`).
263
-
- Use `DROP ALL OBJECTS` via JDBC Execute Query to reset an H2 database during development.
264
+
- The default datasource when using the Frank!Framework is `jdbc/${instance.name.lc}` (e.g., `jdbc/frank2manual`).
265
+
- Use `DROP ALL OBJECTS` via JDBC Execute Query to reset an H2 database during development, this however does not reapply the migration changelog!
264
266
265
267
## XML Validation with XSD
266
268
@@ -319,20 +321,23 @@ Use `<XsltPipe>` with an XSLT stylesheet to transform XML messages:
319
321
```
320
322
321
323
## Transactions
322
-
323
-
Add `transactionAttribute` to the `<Pipeline>` element to wrap all operations in a database transaction.
324
+
Transactions are configured in a `<Receiver>` or in a `<Pipeline>`. You can control the use of transactions via the `transactionAttribute` on either of the elements to wrap all operations in a database transaction. If configured on the `Receiver` and not the `Pipeline`, the `Pipeline` will inherit the given transaction controlled by the `Receiver`.
0 commit comments