1
0
Fork 0
mirror of https://github.com/vbatts/git-validation.git synced 2024-12-26 23:46:30 +00:00

Merge pull request #31 from vbatts/valued_rules

add a message_regexp rule
This commit is contained in:
Vincent Batts 2017-10-12 09:26:06 -04:00 committed by GitHub
commit 90191a015b
8 changed files with 205 additions and 22 deletions

View file

@ -30,7 +30,9 @@ Usage of git-validation:
The entire default rule set is run by default:
```console
vbatts@valse ~/src/vb/git-validation (master) $ git-validation -list-rules
"dangling-whitespace" -- checking the presence of dangling whitespaces on line endings
"DCO" -- makes sure the commits are signed
"message_regexp" -- checks the commit message for a user provided regular expression
"short-subject" -- commit subjects are strictly less than 90 (github ellipsis length)
```
@ -92,6 +94,7 @@ vbatts@valse ~/src/vb/git-validation (master) $ GIT_CHECK_EXCLUDE="./vendor" git
```
using the `GIT_CHECK_EXCLUDE` environment variable
## Rules
Default rules are added by registering them to the `validate` package.

17
main.go
View file

@ -9,6 +9,7 @@ import (
_ "github.com/vbatts/git-validation/rules/danglingwhitespace"
_ "github.com/vbatts/git-validation/rules/dco"
_ "github.com/vbatts/git-validation/rules/messageregexp"
_ "github.com/vbatts/git-validation/rules/shortsubject"
"github.com/vbatts/git-validation/validate"
)
@ -47,10 +48,20 @@ func main() {
return
}
// reduce the set being run
rules := validate.RegisteredRules
// rules to be used
var rules []validate.Rule
for _, r := range validate.RegisteredRules {
// only those that are Default
if r.Default {
rules = append(rules, r)
}
}
// or reduce the set being run to what the user provided
if *flRun != "" {
rules = validate.FilterRules(rules, validate.SanitizeFilters(*flRun))
rules = validate.FilterRules(validate.RegisteredRules, validate.SanitizeFilters(*flRun))
}
if os.Getenv("DEBUG") != "" {
log.Printf("%#v", rules) // XXX maybe reduce this list
}
var commitRange = *flCommitRange

View file

@ -12,6 +12,7 @@ var (
Name: "dangling-whitespace",
Description: "checking the presence of dangling whitespaces on line endings",
Run: ValidateDanglingWhitespace,
Default: true,
}
)
@ -20,7 +21,7 @@ func init() {
}
// ValidateDanglingWhitespace runs Git's check to look for whitespace errors.
func ValidateDanglingWhitespace(c git.CommitEntry) (vr validate.Result) {
func ValidateDanglingWhitespace(r validate.Rule, c git.CommitEntry) (vr validate.Result) {
vr.CommitEntry = c
vr.Msg = "commit does not have any whitespace errors"
vr.Pass = true

View file

@ -20,11 +20,12 @@ var (
Name: "DCO",
Description: "makes sure the commits are signed",
Run: ValidateDCO,
Default: true,
}
)
// ValidateDCO checks that the commit has been signed off, per the DCO process
func ValidateDCO(c git.CommitEntry) (vr validate.Result) {
func ValidateDCO(r validate.Rule, c git.CommitEntry) (vr validate.Result) {
vr.CommitEntry = c
if len(strings.Split(c["parent"], " ")) > 1 {
vr.Pass = true

View file

@ -0,0 +1,61 @@
package messageregexp
import (
"fmt"
"regexp"
"strings"
"github.com/vbatts/git-validation/git"
"github.com/vbatts/git-validation/validate"
)
func init() {
validate.RegisterRule(RegexpRule)
}
var (
// RegexpRule for validating a user provided regex on the commit messages
RegexpRule = validate.Rule{
Name: "message_regexp",
Description: "checks the commit message for a user provided regular expression",
Run: ValidateMessageRegexp,
Default: false, // only for users specifically calling it through -run ...
}
)
// ValidateMessageRegexp is the message regex func to run
func ValidateMessageRegexp(r validate.Rule, c git.CommitEntry) (vr validate.Result) {
if r.Value == "" {
vr.Pass = true
vr.Msg = "noop: message_regexp value is blank"
return vr
}
re := regexp.MustCompile(r.Value)
vr.CommitEntry = c
if len(strings.Split(c["parent"], " ")) > 1 {
vr.Pass = true
vr.Msg = "merge commits are not checked for message_regexp"
return vr
}
hasValid := false
for _, line := range strings.Split(c["subject"], "\n") {
if re.MatchString(line) {
hasValid = true
}
}
for _, line := range strings.Split(c["body"], "\n") {
if re.MatchString(line) {
hasValid = true
}
}
if !hasValid {
vr.Pass = false
vr.Msg = fmt.Sprintf("commit message does not match %q", r.Value)
} else {
vr.Pass = true
vr.Msg = fmt.Sprintf("commit message matches %q", r.Value)
}
return vr
}

View file

@ -11,6 +11,7 @@ var (
Name: "short-subject",
Description: "commit subjects are strictly less than 90 (github ellipsis length)",
Run: ValidateShortSubject,
Default: true,
}
)
@ -20,7 +21,7 @@ func init() {
// ValidateShortSubject checks that the commit's subject is strictly less than
// 90 characters (preferably not more than 72 chars).
func ValidateShortSubject(c git.CommitEntry) (vr validate.Result) {
func ValidateShortSubject(r validate.Rule, c git.CommitEntry) (vr validate.Result) {
if len(c["subject"]) >= 90 {
vr.Pass = false
vr.Msg = "commit subject exceeds 90 characters"

View file

@ -1,33 +1,40 @@
package validate
import (
"sort"
"strings"
"sync"
"github.com/vbatts/git-validation/git"
)
var (
// RegisteredRules are the standard validation to perform on git commits
RegisteredRules = []Rule{}
// RegisteredRules are the avaible validation to perform on git commits
RegisteredRules = []Rule{}
registerRuleLock = sync.Mutex{}
)
// RegisterRule includes the Rule in the avaible set to use
func RegisterRule(vr Rule) {
registerRuleLock.Lock()
defer registerRuleLock.Unlock()
RegisteredRules = append(RegisteredRules, vr)
}
// Rule will operate over a provided git.CommitEntry, and return a result.
type Rule struct {
Name string // short name for reference in in the `-run=...` flag
Value string // value to configure for the rule (i.e. a regexp to check for in the commit message)
Description string // longer Description for readability
Run func(git.CommitEntry) Result
Run func(Rule, git.CommitEntry) Result
Default bool // whether the registered rule is run by default
}
// Commit processes the given rules on the provided commit, and returns the result set.
func Commit(c git.CommitEntry, rules []Rule) Results {
results := Results{}
for _, r := range rules {
results = append(results, r.Run(c))
results = append(results, r.Run(r, c))
}
return results
}
@ -54,28 +61,74 @@ func (vr Results) PassFail() (pass int, fail int) {
return pass, fail
}
// SanitizeFilters takes a comma delimited list and returns the cleaned items in the list
func SanitizeFilters(filt string) (excludes []string) {
for _, item := range strings.Split(filt, ",") {
excludes = append(excludes, strings.TrimSpace(item))
// SanitizeFilters takes a comma delimited list and returns the trimmend and
// split (on ",") items in the list
func SanitizeFilters(filtStr string) (filters []string) {
for _, item := range strings.Split(filtStr, ",") {
filters = append(filters, strings.TrimSpace(item))
}
return
}
// FilterRules takes a set of rules and a list of short names to exclude, and returns the reduced set.
// The comparison is case insensitive.
func FilterRules(rules []Rule, excludes []string) []Rule {
// FilterRules takes a set of rules and a list of short names to include, and
// returns the reduced set. The comparison is case insensitive.
//
// Some `includes` rules have values assigned to them.
// i.e. -run "dco,message_regexp='^JIRA-[0-9]+ [A-Z].*$'"
//
func FilterRules(rules []Rule, includes []string) []Rule {
ret := []Rule{}
for _, r := range rules {
for _, e := range excludes {
if strings.ToLower(r.Name) == strings.ToLower(e) {
ret = append(ret, r)
for i := range includes {
if strings.Contains(includes[i], "=") {
chunks := strings.SplitN(includes[i], "=", 2)
if strings.ToLower(r.Name) == strings.ToLower(chunks[0]) {
// for these rules, the Name won't be unique per se. There may be
// multiple "regexp=" with different values. We'll need to set the
// .Value = chunk[1] and ensure r is dup'ed so they don't clobber
// each other.
newR := Rule(r)
newR.Value = chunks[1]
ret = append(ret, newR)
}
} else {
if strings.ToLower(r.Name) == strings.ToLower(includes[i]) {
ret = append(ret, r)
}
}
}
}
return ret
}
// StringsSliceEqual compares two string arrays for equality
func StringsSliceEqual(a, b []string) bool {
if !sort.StringsAreSorted(a) {
sort.Strings(a)
}
if !sort.StringsAreSorted(b) {
sort.Strings(b)
}
for i := range b {
if !StringsSliceContains(a, b[i]) {
return false
}
}
for i := range a {
if !StringsSliceContains(b, a[i]) {
return false
}
}
return true
}
// StringsSliceContains checks for the presence of a word in string array
func StringsSliceContains(a []string, b string) bool {
if !sort.StringsAreSorted(a) {
sort.Strings(a)
}
i := sort.SearchStrings(a, b)
return i < len(a) && a[i] == b
}

52
validate/rules_test.go Normal file
View file

@ -0,0 +1,52 @@
package validate
import (
"testing"
)
func TestSanitizeRules(t *testing.T) {
set := []struct {
input string
output []string
}{
{
input: "apples, oranges , bananas",
output: []string{"apples", "oranges", "bananas"},
},
{
input: "apples, oranges , bananas, peaches='with cream'",
output: []string{"apples", "oranges", "bananas", "peaches='with cream'"},
},
}
for i := range set {
filt := SanitizeFilters(set[i].input)
if !StringsSliceEqual(filt, set[i].output) {
t.Errorf("expected output like %v, but got %v", set[i].output, filt)
}
}
}
func TestSliceHelpers(t *testing.T) {
set := []struct {
A, B []string
Equal bool
}{
{
A: []string{"apples", "bananas", "oranges", "mango"},
B: []string{"oranges", "bananas", "apples", "mango"},
Equal: true,
},
{
A: []string{"apples", "bananas", "oranges", "mango"},
B: []string{"waffles"},
Equal: false,
},
}
for i := range set {
got := StringsSliceEqual(set[i].A, set[i].B)
if got != set[i].Equal {
t.Errorf("expected %d A and B comparison to be %t, but got %t", i, set[i].Equal, got)
}
}
}