Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
235 changes: 117 additions & 118 deletions core/connectors/sources/postgres_source/src/lib.rs
Original file line number Diff line number Diff line change
Expand Up @@ -838,7 +838,7 @@ impl PostgresSource {
continue;
}

let value = self.extract_column_value(row, i)?;
let value = extract_column_value(row, i)?;
data.insert(column_name.clone(), value.clone());

if column.name() == config.tracking_column {
Expand Down Expand Up @@ -933,124 +933,123 @@ impl PostgresSource {
}
}
}
}

fn extract_column_value(
&self,
row: &sqlx::postgres::PgRow,
column_index: usize,
) -> Result<serde_json::Value, Error> {
let column = &row.columns()[column_index];
let type_name = column.type_info().name();

match type_name {
"BOOL" => {
let value: Option<bool> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value
.map(serde_json::Value::Bool)
.unwrap_or(serde_json::Value::Null))
}
"INT2" => {
let value: Option<i16> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value
.map(|v| serde_json::Value::from(v as i64))
.unwrap_or(serde_json::Value::Null))
}
"INT4" => {
let value: Option<i32> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value
.map(|v| serde_json::Value::from(v as i64))
.unwrap_or(serde_json::Value::Null))
}
"INT8" => {
let value: Option<i64> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value
.map(serde_json::Value::from)
.unwrap_or(serde_json::Value::Null))
}
"FLOAT4" => {
let value: Option<f32> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value
.map(|v| serde_json::Value::from(v as f64))
.unwrap_or(serde_json::Value::Null))
}
"FLOAT8" => {
let value: Option<f64> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value
.map(serde_json::Value::from)
.unwrap_or(serde_json::Value::Null))
}
"NUMERIC" => {
let value: Option<String> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value
.and_then(|s| s.parse::<f64>().ok())
.map(serde_json::Value::from)
.unwrap_or(serde_json::Value::Null))
}
"VARCHAR" | "TEXT" | "CHAR" => {
let value: Option<String> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value
.map(serde_json::Value::String)
.unwrap_or(serde_json::Value::Null))
}
"TIMESTAMP" | "TIMESTAMPTZ" => {
let value: Option<DateTime<Utc>> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value
.map(|dt| serde_json::Value::String(dt.to_rfc3339()))
.unwrap_or(serde_json::Value::Null))
}
"UUID" => {
let value: Option<Uuid> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value
.map(|u| serde_json::Value::String(u.to_string()))
.unwrap_or(serde_json::Value::Null))
}
"JSON" | "JSONB" => {
let value: Option<serde_json::Value> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value.unwrap_or(serde_json::Value::Null))
}
"BYTEA" => {
let value: Option<Vec<u8>> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value
.map(|bytes| {
use base64::Engine;
serde_json::Value::String(
base64::engine::general_purpose::STANDARD.encode(&bytes),
)
})
.unwrap_or(serde_json::Value::Null))
}
_ => {
let value: Option<String> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value
.map(serde_json::Value::String)
.unwrap_or(serde_json::Value::Null))
}
fn extract_column_value(
row: &sqlx::postgres::PgRow,
column_index: usize,
) -> Result<serde_json::Value, Error> {
let column = &row.columns()[column_index];
let type_name = column.type_info().name();

match type_name {
"BOOL" => {
let value: Option<bool> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value
.map(serde_json::Value::Bool)
.unwrap_or(serde_json::Value::Null))
}
"INT2" => {
let value: Option<i16> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value
.map(|v| serde_json::Value::from(v as i64))
.unwrap_or(serde_json::Value::Null))
}
"INT4" => {
let value: Option<i32> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value
.map(|v| serde_json::Value::from(v as i64))
.unwrap_or(serde_json::Value::Null))
}
"INT8" => {
let value: Option<i64> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value
.map(serde_json::Value::from)
.unwrap_or(serde_json::Value::Null))
}
"FLOAT4" => {
let value: Option<f32> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value
.map(|v| serde_json::Value::from(v as f64))
.unwrap_or(serde_json::Value::Null))
}
"FLOAT8" => {
let value: Option<f64> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value
.map(serde_json::Value::from)
.unwrap_or(serde_json::Value::Null))
}
"NUMERIC" => {
let value: Option<String> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value
.and_then(|s| s.parse::<f64>().ok())
.map(serde_json::Value::from)
.unwrap_or(serde_json::Value::Null))
}
"VARCHAR" | "TEXT" | "CHAR" | "BPCHAR" => {
Comment thread
hubcio marked this conversation as resolved.
let value: Option<String> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value
.map(serde_json::Value::String)
.unwrap_or(serde_json::Value::Null))
}
"TIMESTAMP" | "TIMESTAMPTZ" => {
let value: Option<DateTime<Utc>> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value
.map(|dt| serde_json::Value::String(dt.to_rfc3339()))
.unwrap_or(serde_json::Value::Null))
}
"UUID" => {
let value: Option<Uuid> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value
.map(|u| serde_json::Value::String(u.to_string()))
.unwrap_or(serde_json::Value::Null))
}
"JSON" | "JSONB" => {
let value: Option<serde_json::Value> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value.unwrap_or(serde_json::Value::Null))
}
"BYTEA" => {
let value: Option<Vec<u8>> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value
.map(|bytes| {
use base64::Engine;
serde_json::Value::String(
base64::engine::general_purpose::STANDARD.encode(&bytes),
)
})
.unwrap_or(serde_json::Value::Null))
}
_ => {
let value: Option<String> = row
.try_get(column_index)
.map_err(|_| Error::InvalidRecord)?;
Ok(value
.map(serde_json::Value::String)
.unwrap_or(serde_json::Value::Null))
}
}
}
Expand Down
7 changes: 5 additions & 2 deletions core/integration/tests/connectors/fixtures/postgres/source.rs
Original file line number Diff line number Diff line change
Expand Up @@ -60,7 +60,8 @@ impl PostgresSourceJsonFixture {
count INTEGER NOT NULL,
amount DOUBLE PRECISION NOT NULL,
active BOOLEAN NOT NULL,
timestamp BIGINT NOT NULL
timestamp BIGINT NOT NULL,
tag CHAR(10) NOT NULL
)",
Self::TABLE
);
Expand All @@ -81,8 +82,9 @@ impl PostgresSourceJsonFixture {
active: bool,
timestamp: i64,
) {
let tag = format!("{:<10}", format!("tag_{id}"));
let query = format!(
"INSERT INTO {} (id, name, count, amount, active, timestamp) VALUES ($1, $2, $3, $4, $5, $6)",
"INSERT INTO {} (id, name, count, amount, active, timestamp, tag) VALUES ($1, $2, $3, $4, $5, $6, $7)",
Self::TABLE
);
sqlx::query(&query)
Expand All @@ -92,6 +94,7 @@ impl PostgresSourceJsonFixture {
.bind(amount)
.bind(active)
.bind(timestamp)
.bind(&tag)
.execute(pool)
.await
.unwrap_or_else(|e| panic!("Failed to insert row: {e}"));
Expand Down
17 changes: 17 additions & 0 deletions core/integration/tests/connectors/postgres/postgres_source.rs
Original file line number Diff line number Diff line change
Expand Up @@ -64,6 +64,7 @@ async fn json_rows_source_produces_messages_to_iggy(
let consumer_id: Identifier = "test_consumer".try_into().unwrap();

let mut received: Vec<DatabaseRecord> = Vec::new();
let mut raw_payloads: Vec<Vec<u8>> = Vec::new();
for _ in 0..POLL_ATTEMPTS {
if let Ok(polled) = client
.poll_messages(
Expand All @@ -79,6 +80,7 @@ async fn json_rows_source_produces_messages_to_iggy(
{
for msg in polled.messages {
if let Ok(record) = serde_json::from_slice(&msg.payload) {
raw_payloads.push(msg.payload.to_vec());
received.push(record);
}
}
Expand Down Expand Up @@ -110,6 +112,21 @@ async fn json_rows_source_produces_messages_to_iggy(
"Message data mismatch at record {i}"
);
}

// Verify BPCHAR (CHAR(n)) column extraction — Postgres reports CHAR(n) as BPCHAR
for (i, raw) in raw_payloads.iter().enumerate() {
let json: serde_json::Value =
serde_json::from_slice(raw).expect("Failed to parse raw payload");
let tag = json["data"]["tag"]
.as_str()
.unwrap_or_else(|| panic!("Missing BPCHAR 'tag' field in record {i}"));
let expected_tag = format!("tag_{}", test_messages[i].id);
assert_eq!(
tag.trim(),
expected_tag,
"BPCHAR tag mismatch at record {i}"
);
}
}

#[iggy_harness(
Expand Down
Loading