Initial commit

Signed-off-by: Byron Ruth <b@devel.io>

.gitignore

@@ -0,0 +1,3 @@
+*.sw?
+vendor/
+dist/

Makefile

@@ -0,0 +1,49 @@
+PROG_NAME := "diff-table"
+IMAGE_NAME := "dbhi/diff-table"
+CMD_PATH := "."
+
+GIT_SHA := $(shell git log -1 --pretty=format:"%h" .)
+GIT_TAG := $(shell git describe --tags --exact-match . 2>/dev/null)
+GIT_BRANCH := $(shell git symbolic-ref -q --short HEAD)
+GIT_VERSION := $(shell git log -1 --pretty=format:"%h (%ci)" .)
+
+build:
+	go build -ldflags "-X \"main.buildVersion=$(GIT_VERSION)\"" \
+		-o $(GOPATH)/bin/$(PROG_NAME) $(CMD_PATH)
+
+dist-build:
+	mkdir -p dist
+
+	gox -output="./dist/{{.OS}}-{{.Arch}}/$(PROG_NAME)" \
+		-ldflags "-X \"main.buildVersion=$(GIT_VERSION)\"" \
+		-os "windows linux darwin" \
+		-arch "amd64" $(CMD_PATH) > /dev/null
+
+dist-zip:
+	cd dist && zip $(PROG_NAME)-darwin-amd64.zip darwin-amd64/*
+	cd dist && zip $(PROG_NAME)-linux-amd64.zip linux-amd64/*
+	cd dist && zip $(PROG_NAME)-windows-amd64.zip windows-amd64/*
+
+dist: dist-build dist-zip
+
+docker:
+	docker build -t ${IMAGE_NAME}:${GIT_SHA} .
+	docker tag ${IMAGE_NAME}:${GIT_SHA} ${IMAGE_NAME}:${GIT_BRANCH}
+	if [ -n "${GIT_TAG}" ] ; then \
+		docker tag ${IMAGE_NAME}:${GIT_SHA} ${IMAGE_NAME}:${GIT_TAG} ; \
+	fi;
+	if [ "${GIT_BRANCH}" == "master" ]; then \
+		docker tag ${IMAGE_NAME}:${GIT_SHA} ${IMAGE_NAME}:latest ; \
+	fi;
+
+docker-push:
+	docker push ${IMAGE_NAME}:${GIT_SHA}
+	docker push ${IMAGE_NAME}:${GIT_BRANCH}
+	if [ -n "${GIT_TAG}" ]; then \
+		docker push ${IMAGE_NAME}:${GIT_TAG} ; \
+	fi;
+	if [ "${GIT_BRANCH}" == "master" ]; then \
+		docker push ${IMAGE_NAME}:latest ; \
+	fi;
+
+.PHONY: build dist-build dist

README.md

@@ -0,0 +1,88 @@
+# diff-table
+
+A tool to compare two tables of data. Currently the tool only supports executing queries against a Postgres database, but there are plans to add more SQL drivers and flat file support.
+
+The primary use case is to compare the output of a query executed at different points in time. For example, in a batch ETL process that runs every night, you can compare the previous batch with the new batch.
+
+## Install
+
+Download a [release](https://github.com/chop-dbhi/diff-table/releases).
+
+## Usage
+
+This is a minimum example which shows the required options.
+
+```
+diff-table \
+  -db postgres://localhost:5432/postgres \
+  -table1 data_v1 \
+  -table2 data_v2 \
+  -key id
+```
+
+Here are the full set of options.
+
+```
+Usage of diff-table:
+  -db string
+    	Database 1 connection URL.
+  -db2 string
+    	Database 2 connection URL. Defaults to db option.
+  -diff
+    	Diff row values and output new rows and changes.
+  -key string
+    	Required comma-separate list of columns.
+  -schema string
+    	Name of the first schema.
+  -schema2 string
+    	Name of the second schema. Defaults to schema option.
+  -table1 string
+    	Name of the first table.
+  -table2 string
+    	Name of the second table.
+```
+
+## Exampl
+
+```
+diff-table \
+  -db postgres://localhost:5432/postgres \
+  -table1 data_v1 \
+  -table2 data_v2 \
+  -key id
+```
+
+The output is a JSON encoded value which various information about the table differences. If `-diff` is supplied, the `row_diffs` and `new_rows` are tracked and outputed as well. Note, this may significantly increase memory usage if the tables are vastly different.
+
+```
+{
+  "total_rows": 2055856,
+  "columns_added": [],
+  "columns_dropped": [],
+  "type_changes": {},
+  "rows_added": 1,
+  "rows_deleted": 0,
+  "rows_changed": 1,
+  "row_diffs": [
+    {
+      "key": {
+        "id": 2009
+      },
+      "changes": {
+        "val": {
+          "old": 0.7576323747634888,
+          "new": 1.323199987411499
+        }
+      }
+    }
+  ],
+  "new_rows": [
+    {
+      "id": 2010,
+      "val": 1.53921932383223
+    }
+  ]
+}
+```
+
+

lock.json

@@ -0,0 +1,13 @@
+{
+    "memo": "edf2341b8b06db87de490e562dc95fe9bd9466b2b5dadcf6ed1daf3c99a3f046",
+    "projects": [
+        {
+            "name": "github.com/lib/pq",
+            "revision": "0dad96c0b94f8dee039aa40467f767467392a0af",
+            "packages": [
+                ".",
+                "oid"
+            ]
+        }
+    ]
+}