Export Hooks: Glue Catalog exporter

pkg/actions/lua/lakefs/catalogexport/glue_exporter.lua

@@ -0,0 +1,148 @@
+local pathlib = require("path")
+local json = require("encoding/json")
+local lakefs = require("lakefs")
+local extractor = require("lakefs/catalogexport/table_extractor")
+local utils = require("lakefs/catalogexport/internal")
+local sym_exporter = require("lakefs/catalogexport/symlink_exporter")
+
+--[[
+    Generate glue table name
+    @descriptor(Table): object from (i.e _lakefs_tables/my_table.yaml)
+    @action_info(Table): the global action object 
+]]
+local function get_full_table_name(descriptor, action_info)
+    local commit_id = action_info.commit_id
+    local repo_id = action_info.repository_id
+    local branch_or_tag = utils.ref_from_branch_or_tag(action_info)
+    local sha = utils.short_digest(commit_id)
+    return string.format("%s_%s_%s_%s", descriptor.name, repo_id, branch_or_tag, sha)
+end
+
+-- map hive to glue types
+local typesMapping = {
+    integer = "int"
+}
+
+-- helper function to convert hive col to part of glue create table input
+local function hive_col_to_glue(col)
+    return {
+        Name = col.name,
+        Type = typesMapping[col.type] or col.type,
+        Comment = col.comment,
+        Parameters = col.parameters
+    }
+end
+
+-- Create list of partitions for Glue input from a Hive descriptor 
+local function hive_partitions_to_glue_input(descriptor)
+    local partitions = {}
+    local cols = descriptor.schema.fields or {}
+    -- columns list to map by name
+    for _, c in ipairs(cols) do
+        cols[c.name] = c
+    end
+    -- iterate partitions order and find them in the fields, the order determines the path in storage
+    for _, part_key in ipairs(descriptor.partition_columns) do
+        local col = cols[part_key]
+        if col == nil then
+            error(string.format("partition name `%s` not found in table `%s`", part_key, descriptor.name))
+        end
+        table.insert(partitions, hive_col_to_glue(col))
+    end
+    return partitions
+end
+
+-- Create list of columns for Glue excluding partitions
+local function hive_columns_to_glue_input(descriptor)
+    -- create set of partition names since they must not appear in the columns input in glue 
+    local partition_names = {}
+    for _, p in ipairs(descriptor.partition_columns) do
+        partition_names[p] = true
+    end
+    -- create columns as inputs for glue 
+    local columns = {}
+    local cols = descriptor.schema.fields or {}
+    for _, col in ipairs(cols) do
+        if not partition_names[col.name] then -- not a partition 
+            table.insert(columns, hive_col_to_glue(col))
+        end
+    end
+    return columns
+end
+
+-- default location value (e.g root location of either partitions or flat symlink.txt file)
+local function get_table_location(storage_base_prefix, descriptor, action_info)
+    local commit_id = action_info.commit_id
+    local export_base_uri = sym_exporter.get_storage_uri_prefix(storage_base_prefix, commit_id, action_info)
+    return pathlib.join("/", export_base_uri, descriptor.name)
+end
+
+-- create a standard AWS Glue table input (i.e not Apache Iceberg), add input values to base input and configure the rest
+local function build_glue_create_table_input(base_input, descriptor, symlink_location, columns, partitions, action_info,
+    options)
+    local input = utils.deepcopy(base_input)
+    local opts = options or {}
+    input.Name = opts.table_name or get_full_table_name(descriptor, action_info)
+    input.PartitionKeys = array(partitions)
+    input.TableType = "EXTERNAL_TABLE"
+    input.StorageDescriptor.Columns = array(columns)
+    input.StorageDescriptor.Location = symlink_location
+    return input
+end
+
+--[[
+    create a standard glue table in glue catalog
+    @glue: AWS glue client
+    @db(string): glue database name
+    @table_src_path(string): path to table spec (i.e _lakefs_tables/my_table.yaml)
+    @create_table_input(Table): struct mapping to table_input in AWS https://docs.aws.amazon.com/glue/latest/webapi/API_CreateTable.html#API_CreateTable_RequestSyntax
+    should contain inputs describing the data format (i.e InputFormat, OutputFormat, SerdeInfo) since the exporter is agnostic to this. 
+    by default this function will configure table location and schema.
+    @action_info(Table): the global action object
+    @options:
+    - table_name(string): override default glue table name
+    - debug(boolean)
+    - export_base_uri(string): override the default prefix in S3 for symlink location i.e s3://other-bucket/path/
+]]
+local function export_glue(glue, db, table_src_path, create_table_input, action_info, options)
+    local opts = options or {}
+    local repo_id = action_info.repository_id
+    local commit_id = action_info.commit_id
+
+    -- get table desctiptor from _lakefs_tables/
+    local descriptor = extractor.get_table_descriptor(lakefs, repo_id, commit_id, table_src_path)
+
+    -- get table symlink location uri 
+    local base_prefix = opts.export_base_uri or action_info.storage_namespace
+    local symlink_location = get_table_location(base_prefix, descriptor, action_info)
+
+    -- parse Hive table
+    local columns = {}
+    local partitions = {}
+    if descriptor.type == "hive" then
+        -- convert hive cols/partitions to glue
+        partitions = hive_partitions_to_glue_input(descriptor)
+        columns = hive_columns_to_glue_input(descriptor)
+    else
+        error("table " .. descriptor.type .. " in path " .. table_src_path .. " not supported")
+    end
+
+    -- finallize create glue table input
+    local table_input = build_glue_create_table_input(create_table_input, descriptor, symlink_location, columns,
+        partitions, action_info, opts)
+
+    -- create table
+    local json_input = json.marshal(table_input)
+    if opts.debug then
+        print("Creating Glue Table - input:", json_input)
+    end
+    glue.create_table(db, json_input)
+    return {
+        table_input = table_input
+    }
+end
+
+return {
+    get_full_table_name=get_full_table_name,
+    export_glue = export_glue
+}

pkg/actions/lua/lakefs/catalogexport/hive.lua

@@ -1,7 +1,7 @@
 local pathlib = require("path")
 local utils = require("lakefs/catalogexport/internal")
 local strings = require("strings")
-local DEFAULT_PAGE_SIZE = 30 
+local DEFAULT_PAGE_SIZE = 30
 
 -- extract partition prefix from full path
 local function extract_partitions_path(partitions, path)
@@ -11,7 +11,7 @@ local function extract_partitions_path(partitions, path)
     local idx = 1
     local is_partition_prefix = strings.has_prefix(path, partitions[1])
     for part_idx, partition in ipairs(partitions) do
-        local col_substr = "/" ..  partition .. "="
+        local col_substr = "/" .. partition .. "="
         -- if partition is the path prefix and we are the that first partition remove /
         if part_idx == 1 and is_partition_prefix then
             col_substr = partition .. "="
@@ -20,7 +20,7 @@ local function extract_partitions_path(partitions, path)
         if i == nil then
             return nil
         end
-        local separator_idx = string.find(path, "/", j+1)
+        local separator_idx = string.find(path, "/", j + 1)
         -- verify / found and there is something in between = ... / 
         if separator_idx == nil or separator_idx <= (j + 1) then
             return nil
@@ -33,7 +33,8 @@ end
 -- Hive format partition iterator each result set is a collection of files under the same partition
 local function extract_partition_pager(client, repo_id, commit_id, base_path, partition_cols, page_size)
     local target_partition = ""
-    local pager = utils.lakefs_object_pager(client, repo_id, commit_id, "", base_path,"", page_size or DEFAULT_PAGE_SIZE)
+    local pager = utils.lakefs_object_pager(client, repo_id, commit_id, "", base_path, "",
+        page_size or DEFAULT_PAGE_SIZE)
     local page = pager()
     return function()
         if page == nil then
@@ -66,13 +67,13 @@ local function extract_partition_pager(client, repo_id, commit_id, base_path, pa
                     size = entry.size_bytes,
                     checksum = entry.checksum
                 })
-                -- remove entry only if its part of the current partition
-                table.remove(page, 1)
             end
+            -- remove entry (if its part of current partition, hidden files etc) from the entry set
+            table.remove(page, 1)
         end
     end
 end
 
 return {
-    extract_partition_pager=extract_partition_pager,
-}
+    extract_partition_pager = extract_partition_pager
+}

pkg/actions/lua/lakefs/catalogexport/internal.lua

@@ -1,6 +1,21 @@
 local url = require("net/url")
 local DEFAULT_SHORT_DIGEST_LEN=6
 
+local function deepcopy(orig)
+    local orig_type = type(orig)
+    local copy
+    if orig_type == 'table' then
+        copy = {}
+        for orig_key, orig_value in next, orig, nil do
+            copy[deepcopy(orig_key)] = deepcopy(orig_value)
+        end
+        setmetatable(copy, deepcopy(getmetatable(orig)))
+    else -- number, string, boolean, etc
+        copy = orig
+    end
+    return copy
+end
+
 local function short_digest(digest, len)
     return digest:sub(1, len or DEFAULT_SHORT_DIGEST_LEN)
 end 
@@ -52,6 +67,7 @@ local function parse_storage_uri(uri)
 end
 
 return {
+    deepcopy=deepcopy,
     parse_storage_uri=parse_storage_uri,
     short_digest=short_digest,
     ref_from_branch_or_tag=ref_from_branch_or_tag,

pkg/actions/lua/lakefs/catalogexport/symlink_exporter.lua

@@ -96,13 +96,14 @@ local function export_s3(s3_client, table_src_path, action_info, options)
         if opts.debug then
             print("S3 writing bucket: " .. location.bucket .. " key: " .. key)
         end
-        put_obj(location.bucket, key, symlink.data)
+        put_object(location.bucket, key, symlink.data)
     end
     return {
         location = location
     }
 end
 
 return {
-    export_s3 = export_s3
+    export_s3 = export_s3,
+    get_storage_uri_prefix=get_storage_uri_prefix,
 }

pkg/actions/lua/storage/aws/aws.go

@@ -0,0 +1,19 @@
+package aws
+
+import (
+	"context"
+
+	"github.com/Shopify/go-lua"
+)
+
+func Open(l *lua.State, ctx context.Context) {
+	open := func(l *lua.State) int {
+		lua.NewLibrary(l, []lua.RegistryFunction{
+			{Name: "s3_client", Function: newS3Client(ctx)},
+			{Name: "glue_client", Function: newGlueClient(ctx)},
+		})
+		return 1
+	}
+	lua.Require(l, "aws", open, false)
+	l.Pop(1)
+}