load-csv (#105)

* load-csv

* fix typos

---------

Co-authored-by: Guy Korland <[email protected]>

.wordlist.txt

@@ -12,6 +12,7 @@ CLI
 CMD
 CSC
 CSV
+CSVs
 Cailliau
 Centos
 ColumnType
@@ -82,6 +83,7 @@ RedisGraph
 RedisInsight
 Roi
 Rueian
+Saldana
 SETUSER
 SDK
 SDKs
@@ -95,10 +97,12 @@ Subcommands
 TF
 TTY
 UNIQIE
+URIs
 UUID
 VKEY
 ValueType
 Vijay
+Vin
 acos
 acyclic
 algo
@@ -112,6 +116,7 @@ args
 ascii
 asin
 atan
+birthyear
 boolean
 ceil
 cli
@@ -124,6 +129,7 @@ costProp
 cpp
 createNodeIndex
 crflynn
+csv
 dbms
 dedup
 dockerfile

cypher/algorithms.md

@@ -1,6 +1,6 @@
 ---
 title: "Algorithms"
-nav_order: 19
+nav_order: 20
 description: >
     FalkorDB supported algorithms like BFS.
 parent: "Cypher Language"

cypher/cypher_support.md

@@ -1,6 +1,6 @@
 ---
 title: "Cypher coverage"
-nav_order: 21
+nav_order: 22
 description: >
     FalkorDB implements a subset of the Cypher language, which is growing as development continues.
 parent: "Cypher Language"

cypher/functions.md

@@ -1,6 +1,6 @@
 ---
 title: "Functions"
-nav_order: 17
+nav_order: 18
 description: >
     Information on all supported functions from the Cypher query language
 parent: "Cypher Language"
@@ -364,4 +364,4 @@ The format for a relationship object in JSON is:
   "start": src_node(node),
   "end": dest_node(node)
 }
-```
+```

cypher/indexing.md

@@ -1,6 +1,6 @@
 ---
 title: "Indexing"
-nav_order: 20
+nav_order: 21
 description: >
     FalkorDB supports single-property indexes for node labels and for relationship type. String, numeric, and geospatial data types can be indexed.
 parent: "Cypher Language"

cypher/load_csv.md

@@ -0,0 +1,113 @@
+---
+title: "LOAD CSV"
+nav_order: 17
+description: >
+    LOAD CSV allows a query to access data within a CSV file
+parent: "Cypher Language"
+---
+
+# LOAD CSV
+
+```sh
+LOAD CSV FROM 'file://actors.csv' AS row
+MERGE (a:Actor {name: row[0]})
+```
+
+`LOAD CSV FROM` accepts a string containing the path to a CSV file,
+the file is parsed line by line, the current line is accessible through the 
+variable specified by `AS`. Each parsed value is treated as a `string`, use
+the right conversion functions e.g. `toInteger` to cast a value to its
+appropriate type.
+
+Additional clauses can follow and accesses the `row` variable
+
+## IMPORTING DATA
+
+### Importing local files
+
+FalkorDB defines a data directory ![see configuration](../configuration)
+Under which local CSV files should be stored, all `file://` URIs are resolved
+relatively to that directory.
+
+In the following example we'll load the `actors.csv` file into FalkorDB.
+
+### actors.csv
+
+| Lee Pace       | 1979      | 
+| ---------------|-----------|
+| Vin Diesel     | 1967      |
+| Chris Pratt    | 1979      |
+| Zoe Saldana    | 1978      |
+
+```sh
+LOAD CSV FROM 'file://actors.csv'
+AS row
+MERGE (a:Actor {name: row[0], birth_year: toInteger(row[1])})
+RETURN a.name, a.birth_year
+```
+
+Note that we've used indices e.g. `row[0]` to access the value at the corresponding
+column.
+
+In case the CSV contains a header row e.g.
+
+### actors.csv
+
+| name           | birthyear |
+| ---------------|-----------|
+| Lee Pace       | 1979      | 
+| Vin Diesel     | 1967      |
+| Chris Pratt    | 1979      |
+| Zoe Saldana    | 1978      |
+
+Then we should use the `WITH HEADERS` variation of the `LOAD CSV` clause
+
+```
+LOAD CSV WITH HEADERS FROM 'file://actors.csv'
+AS row
+MERGE (a:Actor {name: row[name], birth_year: toInteger(row[birthyear])})
+RETURN a.name, a.birth_year
+```
+
+Note when a header row exists and `WITH HEADER` is specified the `row` variable
+is no longer an `array` but rather a `map`, accessing the individual elements
+is done via their column name.
+
+
+### Importing data from multiple CSVs
+
+Building on our previous example we'll introduce a second csv file `acted_in.csv`
+which ties actors to movies they've acted in
+
+
+### acted_in.csv
+
+| actor          | movie          |
+| ---------------|----------------|
+| Lee Pace       | The Fall       | 
+| Vin Diesel     | Fast & Furious |
+| Chris Pratt    | Passengers     |
+| Zoe Saldana    | Avatar         |
+
+
+We'll create a new graph connecting actors to the movies they've acted in
+
+Load actors:
+
+```sh
+LOAD CSV WITH HEADER FROM 'file://actors.csv'
+AS row
+MERGE (a:Actor {name:row['name']})
+```
+
+Load movies and create `ACTED_IN` relations:
+
+```sh
+LOAD CSV WITH HEADER FROM 'file://acted_in.csv'
+AS row
+
+MATCH (a:Actor {name: row['actor']})
+MERGE (m:Movie {title: row['movie']})
+MERGE (a)-[:ACTED_IN]->(m)
+```
+

cypher/procedures.md

@@ -1,6 +1,6 @@
 ---
 title: "Procedures"
-nav_order: 18
+nav_order: 19
 description: >
     Procedures calling using with CALL and YIELD.
 parent: "Cypher Language"

cypher/remove.md

@@ -1,6 +1,6 @@
 ---
 title: "REMOVE"
-nav_order: 22
+nav_order: 23
 description: >
     REMOVE is used to remove attributes from node and relationships, in addition to
     removing labels from nodes.