Parselog & Issue start

app/kgpz.go

@@ -85,7 +85,7 @@ func (k *KGPZ) Enrich() error {
 	}
 
 	// INFO: We pass agents by value since we don't want to block the library
-	agents := k.Library.Agents.All()
+	agents := k.Library.Agents.Everything()
 	go func(agents []xmlprovider.Agent) {
 		k.GND.FetchPersons(agents)
 		k.GND.WriteCache(k.Config.GNDPath)
@@ -103,56 +103,43 @@ func (k *KGPZ) Serialize() {
 	k.gmu.Lock()
 	defer k.gmu.Unlock()
 
+	commit := "staticfile"
+	if k.Repo != nil {
+		commit = k.Repo.Commit
+	}
+
 	issues, err := getXMLFiles(filepath.Join(k.Config.FolderPath, ISSUES_DIR))
 	helpers.Assert(err, "Error getting issues")
 
 	pieces, err := getXMLFiles(filepath.Join(k.Config.FolderPath, PIECES_DIR))
 	helpers.Assert(err, "Error getting pieces")
 
-	lib := xmlprovider.NewLibrary(
-		[]string{filepath.Join(k.Config.FolderPath, AGENTS_PATH)},
-		[]string{filepath.Join(k.Config.FolderPath, PLACES_PATH)},
-		[]string{filepath.Join(k.Config.FolderPath, WORKS_PATH)},
-		[]string{filepath.Join(k.Config.FolderPath, CATEGORIES_PATH)},
-		*issues,
-		*pieces)
-
-	lib.Serialize()
-
-	// TODO: is it neccessary to lock here, sice gmu lock prevents concurrent locking of the library?
 	k.lmu.Lock()
 	defer k.lmu.Unlock()
-
 	if k.Library == nil {
+		lib := xmlprovider.NewLibrary(
+			[]string{filepath.Join(k.Config.FolderPath, AGENTS_PATH)},
+			[]string{filepath.Join(k.Config.FolderPath, PLACES_PATH)},
+			[]string{filepath.Join(k.Config.FolderPath, WORKS_PATH)},
+			[]string{filepath.Join(k.Config.FolderPath, CATEGORIES_PATH)},
+			*issues,
+			*pieces)
+
+		lib.Serialize(commit)
+
 		k.Library = lib
-		return
+	} else {
+		// TODO: where to clear the old data?
+		// How to differentiate between deleted data points and stale data points bc of parse errors?
+		k.Library.SetPaths(
+			[]string{filepath.Join(k.Config.FolderPath, AGENTS_PATH)},
+			[]string{filepath.Join(k.Config.FolderPath, PLACES_PATH)},
+			[]string{filepath.Join(k.Config.FolderPath, WORKS_PATH)},
+			[]string{filepath.Join(k.Config.FolderPath, CATEGORIES_PATH)},
+			*issues,
+			*pieces)
+		k.Library.Serialize(commit)
 	}
-
-	if lib.Agents == nil {
-		lib.Agents = k.Library.Agents
-	}
-
-	if lib.Places == nil {
-		lib.Places = k.Library.Places
-	}
-
-	if lib.Works == nil {
-		lib.Works = k.Library.Works
-	}
-
-	if lib.Categories == nil {
-		lib.Categories = k.Library.Categories
-	}
-
-	if lib.Issues == nil {
-		lib.Issues = k.Library.Issues
-	}
-
-	if lib.Pieces == nil {
-		lib.Pieces = k.Library.Pieces
-	}
-
-	k.Library = lib
 }
 
 func (k *KGPZ) IsDebug() bool {

helpers/logging/parselog.go

@@ -0,0 +1,200 @@
+package logging
+
+import (
+	"fmt"
+	"log/slog"
+	"sync"
+)
+
+// WARNING: do not attempt to set this anywhere besides the init function of this module
+var ParseMessages ParseLogger
+
+type XMLEntityType int64
+
+const (
+	Agent XMLEntityType = iota
+	Place
+	Worke
+	Category
+	Issue
+	Piece
+)
+
+type ParseErrorLevel int64
+
+const (
+	Clean ParseErrorLevel = iota
+	ObjectMessage
+	InfoMessage
+	WarningMessage
+	ErrorMessage
+	FatalMessage
+)
+
+type ParseMessage struct {
+	XMLType     XMLEntityType
+	XMLPath     string
+	Object      fmt.Stringer
+	Message     string
+	MessageType ParseErrorLevel
+}
+
+func (pm ParseMessage) String() string {
+	if pm.Object != nil {
+		return fmt.Sprintf("%s: %s\n%s\n%s", pm.XMLType, pm.XMLPath, pm.Object.String(), pm.Message)
+	}
+	return fmt.Sprintf("%s: %s\n%s", pm.XMLType, pm.XMLPath, pm.Message)
+}
+
+type ParseLogger struct {
+	mu           sync.Mutex
+	ParseInfo    chan ParseMessage
+	ParseErrors  chan ParseMessage
+	ParseObjects chan fmt.Stringer
+	messages     []ParseMessage
+	objects      []fmt.Stringer
+	State        ParseErrorLevel
+	subs         []func(ParseMessage)
+}
+
+func init() {
+	ParseMessages = ParseLogger{
+		ParseInfo:   make(chan ParseMessage, 100),
+		ParseErrors: make(chan ParseMessage, 100),
+	}
+	ParseMessages.Start()
+}
+
+func (pl *ParseLogger) Start() {
+	go func() {
+		for {
+			select {
+			case msg, ok := <-pl.ParseObjects:
+				pl.mu.Lock()
+				pl.objects = append(pl.objects, msg)
+				pl.mu.Unlock()
+				if !ok {
+					pl.ParseObjects = nil
+				}
+			case msg, ok := <-pl.ParseInfo:
+				pl.mu.Lock()
+				pl.messages = append(pl.messages, msg)
+				pl.setState(InfoMessage)
+				pl.mu.Unlock()
+				if !ok {
+					pl.ParseInfo = nil
+				}
+			case msg, ok := <-pl.ParseErrors:
+				pl.mu.Lock()
+				pl.messages = append(pl.messages, msg)
+				pl.setState(msg.MessageType)
+				pl.mu.Unlock()
+				if !ok {
+					pl.ParseErrors = nil
+				}
+			}
+
+			if pl.ParseInfo == nil && pl.ParseInfo == nil && pl.ParseObjects == nil {
+				break
+			}
+		}
+	}()
+}
+
+func (pl *ParseLogger) GetMessages() []ParseMessage {
+	res := make([]ParseMessage, len(pl.messages))
+	pl.mu.Lock()
+	defer pl.mu.Unlock()
+	copy(res, pl.messages)
+	return res
+}
+
+func (pl *ParseLogger) ClearMessages() {
+	pl.mu.Lock()
+	defer pl.mu.Unlock()
+	pl.State = Clean
+	pl.messages = []ParseMessage{}
+}
+
+func (pl *ParseLogger) LogInfo(xmlType XMLEntityType, xmlPath string, object fmt.Stringer, message string) {
+	pl.ParseInfo <- ParseMessage{
+		XMLType:     xmlType,
+		XMLPath:     xmlPath,
+		Object:      object,
+		Message:     message,
+		MessageType: InfoMessage,
+	}
+}
+
+func (pl *ParseLogger) LogError(xmlType XMLEntityType, xmlPath string, object fmt.Stringer, message string) {
+	pl.ParseErrors <- ParseMessage{
+		XMLType:     xmlType,
+		XMLPath:     xmlPath,
+		Object:      object,
+		Message:     message,
+		MessageType: ErrorMessage,
+	}
+}
+
+func (pl *ParseLogger) LogWarning(xmlType XMLEntityType, xmlPath string, object fmt.Stringer, message string) {
+	pl.ParseErrors <- ParseMessage{
+		XMLType:     xmlType,
+		XMLPath:     xmlPath,
+		Object:      object,
+		Message:     message,
+		MessageType: WarningMessage,
+	}
+}
+
+func (pl *ParseLogger) LogFatal(xmlType XMLEntityType, xmlPath string, object fmt.Stringer, message string) {
+	pl.ParseErrors <- ParseMessage{
+		XMLType: xmlType,
+		XMLPath: xmlPath,
+		Object:  object,
+		Message: message,
+	}
+}
+
+func (pl *ParseLogger) setState(state ParseErrorLevel) {
+	if state > pl.State {
+		pl.State = state
+	}
+}
+
+func (pl *ParseLogger) GetState() ParseErrorLevel {
+	return pl.State
+}
+
+func (pl *ParseLogger) Len() int {
+	pl.mu.Lock()
+	defer pl.mu.Unlock()
+	return len(pl.messages)
+}
+
+func (pl *ParseLogger) Subscribe(fn func(ParseMessage)) {
+	pl.mu.Lock()
+	defer pl.mu.Unlock()
+	pl.subs = append(pl.subs, fn)
+}
+
+func (pl *ParseLogger) ResetSubscriptions() {
+	pl.mu.Lock()
+	defer pl.mu.Unlock()
+	pl.subs = []func(ParseMessage){}
+}
+
+func (pl *ParseLogger) PrintObjects() {
+	pl.mu.Lock()
+	defer pl.mu.Unlock()
+	for _, o := range pl.objects {
+		ObjDebug(&o, "Object")
+	}
+}
+
+func (pl *ParseLogger) PrintMessages() {
+	pl.mu.Lock()
+	defer pl.mu.Unlock()
+	for _, m := range pl.messages {
+		slog.Debug(m.String())
+	}
+}

providers/xmlprovider/agents.go

@@ -14,6 +14,7 @@ type Agent struct {
 	Org      bool     `xml:"org,attr"`
 	Identifier
 	AnnotationNote
+	SerializedItem
 }
 
 func (a Agent) String() string {

providers/xmlprovider/categories.go

@@ -11,6 +11,7 @@ type Category struct {
 	SortName string   `xml:"sortiername"`
 	Identifier
 	AnnotationNote
+	SerializedItem
 }
 
 func (c Category) String() string {

providers/xmlprovider/issues.go

@@ -15,6 +15,7 @@ type Issue struct {
 	Additionals []Additional `xml:"beilage"`
 	Identifier
 	AnnotationNote
+	SerializedItem
 }
 
 type Nummer struct {

providers/xmlprovider/item.go

@@ -0,0 +1,19 @@
+package xmlprovider
+
+type SerializedItem struct {
+	Source string
+	Date   string
+	Commit string
+}
+
+func (si SerializedItem) SetSource(s string) {
+	si.Source = s
+}
+
+func (si SerializedItem) SetDate(d string) {
+	si.Date = d
+}
+
+func (si SerializedItem) SetCommit(c string) {
+	si.Commit = c
+}

providers/xmlprovider/pieces.go

@@ -21,6 +21,7 @@ type Piece struct {
 	Title         []string        `xml:"titel"`
 	Identifier
 	AnnotationNote
+	SerializedItem
 }
 
 func (p Piece) String() string {
@@ -35,6 +36,43 @@ func (p Piece) GetIDs() []string {
 
 	// TODO: sensible IDs
 	uid := uuid.New()
-	ret = append(ret, uid.String())
+
+	for _, i := range p.IssueRefs {
+		ret = append(ret, i.Datum+"-"+i.Nr+"-"+uid.String())
+	}
+
+	for _, i := range p.AdditionalRef {
+		ret = append(ret, i.Datum+"-"+i.Nr+"-"+uid.String())
+	}
 	return ret
 }
+
+// TODO: We can make this fast depending on which category to look for
+// but we'll have to define rules for every single category (~35 of them)
+func (p Piece) IsCat(k string) bool {
+	for _, c := range p.CategoryRefs {
+		if c.Category == k {
+			return true
+		}
+	}
+
+	for _, c := range p.WorkRefs {
+		if c.Category == k {
+			return true
+		}
+	}
+
+	for _, c := range p.AgentRefs {
+		if c.Category == k {
+			return true
+		}
+	}
+
+	for _, c := range p.PieceRefs {
+		if c.Category == k {
+			return true
+		}
+	}
+
+	return false
+}

providers/xmlprovider/places.go

@@ -12,6 +12,7 @@ type Place struct {
 	Geo      string   `xml:"geonames"`
 	Identifier
 	AnnotationNote
+	SerializedItem
 }
 
 func (p Place) String() string {

providers/xmlprovider/works.go

@@ -13,6 +13,7 @@ type Work struct {
 	Akteur         []AgentRef `xml:"akteur"`
 	Identifier
 	AnnotationNote
+	SerializedItem
 }
 
 type Citation struct {

providers/xmlprovider/xmlprovider.go

@@ -6,6 +6,7 @@ import (
 	"io"
 	"os"
 	"sync"
+	"time"
 
 	"github.com/Theodor-Springmann-Stiftung/kgpz_web/helpers/logging"
 )
@@ -13,13 +14,20 @@ import (
 type XMLItem interface {
 	fmt.Stringer
 	GetIDs() []string
+	SetSource(string)
+	SetDate(string)
+	SetCommit(string)
+}
+
+type Collection[T XMLItem] struct {
+	Collection []T
+	lock       sync.Mutex
 }
 
 type XMLProvider[T XMLItem] struct {
 	Paths []string
+	// INFO: map is type [string]T
 	Items sync.Map
-
-	mu sync.Mutex
 }
 
 type Library struct {
@@ -47,90 +55,87 @@ func NewLibrary(agentpaths, placepaths, workpaths, categorypaths, issuepaths, pi
 	}
 }
 
-func (l *Library) Serialize() {
+func (l *Library) SetPaths(agentpaths, placepaths, workpaths, categorypaths, issuepaths, piecepaths []string) {
+	l.Agents.Paths = agentpaths
+	l.Places.Paths = placepaths
+	l.Works.Paths = workpaths
+	l.Categories.Paths = categorypaths
+	l.Issues.Paths = issuepaths
+	l.Pieces.Paths = piecepaths
+}
+
+func (l *Library) Serialize(commit string) {
 	wg := sync.WaitGroup{}
-	wg.Add(6)
 
-	go func() {
+	for _, path := range l.Places.Paths {
-		defer wg.Done()
+		wg.Add(1)
-		lwg := sync.WaitGroup{}
+		go func() {
-		for _, path := range l.Places.Paths {
+			l.Places.Serialize(NewPlaceRoot(), path, commit)
-			lwg.Add(1)
+			wg.Done()
-			go l.Places.Serialize(NewPlaceRoot(), path, &lwg)
+		}()
-		}
+	}
-		lwg.Wait()
-	}()
 
-	go func() {
+	for _, path := range l.Agents.Paths {
-		defer wg.Done()
+		wg.Add(1)
-		lwg := sync.WaitGroup{}
+		go func() {
-		for _, path := range l.Agents.Paths {
+			l.Agents.Serialize(NewAgentRoot(), path, commit)
-			lwg.Add(1)
+			wg.Done()
-			go l.Agents.Serialize(NewAgentRoot(), path, &lwg)
+		}()
-		}
+	}
-		lwg.Wait()
-	}()
 
-	go func() {
+	for _, path := range l.Categories.Paths {
-		defer wg.Done()
+		wg.Add(1)
-		lwg := sync.WaitGroup{}
+		go func() {
-		for _, path := range l.Categories.Paths {
+			l.Categories.Serialize(NewCategoryRoot(), path, commit)
-			lwg.Add(1)
+			wg.Done()
-			go l.Categories.Serialize(NewCategoryRoot(), path, &lwg)
+		}()
-		}
+	}
-		lwg.Wait()
-	}()
 
-	go func() {
+	for _, path := range l.Works.Paths {
-		defer wg.Done()
+		wg.Add(1)
-		lwg := sync.WaitGroup{}
+		go func() {
-		for _, path := range l.Works.Paths {
+			l.Works.Serialize(NewWorkRoot(), path, commit)
-			lwg.Add(1)
+			wg.Done()
-			go l.Works.Serialize(NewWorkRoot(), path, &lwg)
+		}()
-		}
+	}
-		lwg.Wait()
-	}()
 
-	go func() {
+	for _, path := range l.Issues.Paths {
-		defer wg.Done()
+		wg.Add(1)
-		lwg := sync.WaitGroup{}
+		go func() {
-		for _, path := range l.Issues.Paths {
+			l.Issues.Serialize(NewIssueRoot(), path, commit)
-			lwg.Add(1)
+			wg.Done()
-			go l.Issues.Serialize(NewIssueRoot(), path, &lwg)
+		}()
-		}
+	}
-		lwg.Wait()
-	}()
 
-	go func() {
+	for _, path := range l.Pieces.Paths {
-		defer wg.Done()
+		wg.Add(1)
-		lwg := sync.WaitGroup{}
+		go func() {
-		for _, path := range l.Pieces.Paths {
+			l.Pieces.Serialize(NewPieceRoot(), path, commit)
-			lwg.Add(1)
+			wg.Done()
-			go l.Pieces.Serialize(NewPieceRoot(), path, &lwg)
+		}()
-		}
+	}
-		lwg.Wait()
-	}()
 
 	wg.Wait()
 }
 
-func (p *XMLProvider[T]) Serialize(dataholder XMLRootElement[T], path string, wg *sync.WaitGroup) error {
+func (p *XMLProvider[T]) Serialize(dataholder XMLRootElement[T], path, commit string) error {
+	date := time.Now().Format("2006-01-02")
 	// Introduce goroutine for every path, locking on append:
 	if err := UnmarshalFile(path, dataholder); err != nil {
 		logging.Error(err, "Could not unmarshal file: "+path)
+		logging.ParseMessages.ParseErrors <- logging.ParseMessage{MessageType: logging.ErrorMessage, Message: "Could not unmarshal file: " + path}
 		return err
 	}
 	for _, item := range dataholder.Children() {
+		item.SetSource(path)
+		item.SetDate(date)
+		item.SetCommit(commit)
 		// INFO: Mostly it's just one ID, so the double loop is not that bad.
 		for _, id := range item.GetIDs() {
 			p.Items.Store(id, item)
 		}
 	}
 
-	if wg != nil {
-		wg.Done()
-	}
-
 	return nil
 }
 
@@ -199,7 +204,9 @@ func (p *XMLProvider[T]) FindKey(fn func(string) bool) []T {
 	return items
 }
 
-func (p *XMLProvider[T]) All() []T {
+// INFO: Do not use this, except when iterating over a collection multiple times (three times or more).
+// Maps are slow to iterate, but many of the Iterations can only be done once.
+func (p *XMLProvider[T]) Everything() []T {
 	var items []T
 	p.Items.Range(func(key, value interface{}) bool {
 		items = append(items, value.(T))

templating/engine.go

@@ -40,6 +40,14 @@ func (e *Engine) MapFuncs(app *app.KGPZ) error {
 	e.AddFunc("GetDate", functions.GetDate)
 	e.AddFunc("MonthName", functions.MonthName)
 	e.AddFunc("MonthNameShort", functions.MonthNameShort)
+
+	e.AddFunc("GetAgent", app.Library.Agents.Item)
+	e.AddFunc("GetPlace", app.Library.Places.Item)
+	e.AddFunc("GetWork", app.Library.Works.Item)
+	e.AddFunc("GetCategory", app.Library.Categories.Item)
+	e.AddFunc("GetIssue", app.Library.Issues.Item)
+	e.AddFunc("GetPiece", app.Library.Pieces.Item)
+
 	return nil
 }
 

viewmodels/singleissue.go

@@ -2,6 +2,7 @@ package viewmodels
 
 import (
 	"strconv"
+	"strings"
 
 	"github.com/Theodor-Springmann-Stiftung/kgpz_web/helpers/logging"
 	"github.com/Theodor-Springmann-Stiftung/kgpz_web/providers/xmlprovider"
@@ -24,30 +25,19 @@ func NewSingleIssueView(y string, No string, lib *xmlprovider.Library) (*SingleI
 	}
 
 	sivm := SingleIssueViewModel{IssueViewModel: *ivm}
-	logging.Info(strconv.Itoa(len(lib.Pieces.All())) + "pieces in library")
+	logging.Info(strconv.Itoa(len(lib.Pieces.Everything())) + "pieces in library")
 
+	lookfor := y + "-" + No
 	lib.Pieces.Items.Range(func(key, value interface{}) bool {
-		a := value.(xmlprovider.Piece)
+		k := key.(string)
-		for _, r := range a.IssueRefs {
+		if strings.HasPrefix(k, lookfor) {
-			if r.Datum == y && r.Nr == No {
+			a := value.(xmlprovider.Piece)
-				p, err := NewPieceView(a)
+			p, err := NewPieceView(a)
-				if err != nil {
+			if err != nil {
-					logging.ObjErr(&a, err)
+				logging.ObjErr(&a, err)
-					continue
+				return true
-				}
-				sivm.Pieces = append(sivm.Pieces, p)
-			}
-		}
-
-		for _, r := range a.AdditionalRef {
-			if r.Datum == y && r.Nr == No {
-				p, err := NewPieceView(a)
-				if err != nil {
-					logging.ObjErr(&a, err)
-					continue
-				}
-				sivm.Additionals = append(sivm.Additionals, p)
 			}
+			sivm.Pieces = append(sivm.Pieces, p)
 		}
 		return true
 	})

viewmodels/year.go

@@ -23,18 +23,19 @@ func YearView(year string, lib *xmlprovider.Library) (*YearViewModel, error) {
 	last := ""
 
 	lib.Issues.Items.Range(func(key, value interface{}) bool {
-		issue := value.(xmlprovider.Issue)
+		k := key.(string)
-		if len(issue.Datum.When) < 4 {
+		if len(k) < 4 {
 			return true
 		}
 
-		date := issue.Datum.When[0:4]
+		date := k[0:4]
 		if date != last {
 			res.PushAvailable(date)
 			last = date
 		}
 
 		if date == year {
+			issue := value.(xmlprovider.Issue)
 			res.PushIssue(issue)
 		}
 		return true

views/routes/issue/body.tmpl

@@ -1 +1,33 @@
-Hello from an issue!
+Issue found!
+
+{{ $model := .model }}
+
+{{ range  $piece := .model.Pieces }}
+	<div>
+		Piece!
+
+		{{ if gt (len $piece.IssueRefs) 1 }}
+			{{ len $piece.IssueRefs }} Teile
+			<ol>
+			{{ range $issue := $piece.IssueRefs }}
+				<li>
+					<a href="/{{- $issue.Datum -}}/{{- $issue.Nr -}}" >
+						{{- $issue.Datum }} Nr. {{ $issue.Nr -}}
+					</a>
+				</li>
+			{{ end }}
+			</ol>
+		{{ end }}
+
+		{{ range $agentref := $piece.AgentRefs }}
+			{{ $agent := GetAgent $agentref.Ref }}
+			{{ if gt (len $agent.Names) 0 }}
+				{{ index $agent.Names 0 }}
+			{{ end }}
+			{{ if (or (eq $agentref.Category "") (eq $agentref.Category "autor")) }}
+				<span>Author</span>
+			{{ end }}
+		{{ end }}
+
+	</div>
+{{ end }}