| 1 | package applog |
| 2 | |
| 3 | import ( |
| 4 | "bufio" |
| 5 | "bytes" |
| 6 | "compress/gzip" |
| 7 | "encoding/json" |
| 8 | "errors" |
| 9 | "fmt" |
| 10 | "io" |
| 11 | "strconv" |
| 12 | "strings" |
| 13 | "time" |
| 14 | |
| 15 | "github.com/carlosframework/platform/internal/release" |
| 16 | ) |
| 17 | |
| 18 | // hourFormat is the UTC hour vocabulary in chunk keys — deliberately the |
| 19 | // same format internal/metrics uses for its raw chunks, so the bucket |
| 20 | // speaks one hour dialect. |
| 21 | const hourFormat = "2006-01-02-15" |
| 22 | |
| 23 | // HourOf formats t as the UTC hour directory a chunk belongs to. |
| 24 | func HourOf(t time.Time) string { return t.UTC().Format(hourFormat) } |
| 25 | |
| 26 | // HourBefore reports whether hour (a HourOf-formatted directory name) ends |
| 27 | // strictly before cutoff's own hour — the retention test. An unparsable |
| 28 | // name never matches: pruning must not delete on a guess. |
| 29 | func HourBefore(hour string, cutoff time.Time) bool { |
| 30 | t, err := time.ParseInLocation(hourFormat, hour, time.UTC) |
| 31 | if err != nil { |
| 32 | return false |
| 33 | } |
| 34 | return t.Add(time.Hour).Sub(cutoff.UTC()) <= 0 && HourOf(cutoff) != hour |
| 35 | } |
| 36 | |
| 37 | // Family names one chunk family under an app — the path segment after |
| 38 | // apps/<sqid>/<app>/. logs/ is the original; errors/ and analytics/ are |
| 39 | // the siblings the errors-and-analytics spec adds (§3.2). The key shape |
| 40 | // below the family segment is identical for every family, which is what |
| 41 | // lets one prune loop, one store-door grant and one reader serve all three. |
| 42 | type Family string |
| 43 | |
| 44 | const ( |
| 45 | FamilyLogs Family = "logs" |
| 46 | FamilyErrors Family = "errors" |
| 47 | FamilyAnalytics Family = "analytics" |
| 48 | ) |
| 49 | |
| 50 | // ErrBadKeySegment is every family builder's refusal: an account or app |
| 51 | // that is not a valid name, or a box or hour that is empty or carries a |
| 52 | // slash. Writers skip the write on it; readers fail the read. A bad |
| 53 | // segment never becomes a key, so no box can write apps/<sqid>//... (an |
| 54 | // app-less route once did exactly that, and hostops.Apps, rightly, refuses |
| 55 | // to walk a bucket holding it). |
| 56 | var ErrBadKeySegment = errors.New("applog: refusing to build a key from an empty or invalid segment") |
| 57 | |
| 58 | // CheckApp is the account and app half of every family builder's check: |
| 59 | // the same names hostops.Apps accepts when it walks apps/. |
| 60 | func CheckApp(acct, app string) error { |
| 61 | if !release.ValidAccount(acct) || !release.ValidApp(app) { |
| 62 | return fmt.Errorf("%w: account %q, app %q", ErrBadKeySegment, acct, app) |
| 63 | } |
| 64 | return nil |
| 65 | } |
| 66 | |
| 67 | // CheckSegment refuses a box, hour or date segment that would collapse or |
| 68 | // split a path level. internal/analytics uses it for its day-file leaf. |
| 69 | func CheckSegment(what, v string) error { |
| 70 | if v == "" || strings.ContainsRune(v, '/') { |
| 71 | return fmt.Errorf("%w: %s %q", ErrBadKeySegment, what, v) |
| 72 | } |
| 73 | return nil |
| 74 | } |
| 75 | |
| 76 | // FamilyPrefix is the app's whole subtree for one family — under |
| 77 | // release.AppPrefix on purpose (spec §3): it is the app's data, so the |
| 78 | // owner gate maps onto the prefix and app trash/purge sweeps it for free. |
| 79 | // An invalid account or app is refused (ErrBadKeySegment), never built. |
| 80 | func FamilyPrefix(fam Family, acct, app string) (string, error) { |
| 81 | if err := CheckApp(acct, app); err != nil { |
| 82 | return "", err |
| 83 | } |
| 84 | return release.AppPrefix(acct, app) + string(fam) + "/", nil |
| 85 | } |
| 86 | |
| 87 | // FamilyBoxPrefix is one box's subtree within a family — the unit of |
| 88 | // write-side ownership: each box prunes only what it wrote. |
| 89 | func FamilyBoxPrefix(fam Family, acct, app, box string) (string, error) { |
| 90 | p, err := FamilyPrefix(fam, acct, app) |
| 91 | if err != nil { |
| 92 | return "", err |
| 93 | } |
| 94 | if err := CheckSegment("box", box); err != nil { |
| 95 | return "", err |
| 96 | } |
| 97 | return p + box + "/", nil |
| 98 | } |
| 99 | |
| 100 | // FamilyHourPrefix is one UTC hour within a box's subtree. |
| 101 | func FamilyHourPrefix(fam Family, acct, app, box, hour string) (string, error) { |
| 102 | p, err := FamilyBoxPrefix(fam, acct, app, box) |
| 103 | if err != nil { |
| 104 | return "", err |
| 105 | } |
| 106 | if err := CheckSegment("hour", hour); err != nil { |
| 107 | return "", err |
| 108 | } |
| 109 | return p + hour + "/", nil |
| 110 | } |
| 111 | |
| 112 | // FamilyChunkKey names one flushed chunk in any family. See ChunkKey for |
| 113 | // why seq and size live in the key. |
| 114 | func FamilyChunkKey(fam Family, acct, app, box, hour string, seq uint64, size int) (string, error) { |
| 115 | p, err := FamilyHourPrefix(fam, acct, app, box, hour) |
| 116 | if err != nil { |
| 117 | return "", err |
| 118 | } |
| 119 | return fmt.Sprintf("%s%016d-%d.jsonl.gz", p, seq, size), nil |
| 120 | } |
| 121 | |
| 122 | // Prefix is the app's whole log subtree: FamilyPrefix for logs. |
| 123 | func Prefix(acct, app string) (string, error) { return FamilyPrefix(FamilyLogs, acct, app) } |
| 124 | |
| 125 | // BoxPrefix is one box's subtree within an app's logs. |
| 126 | func BoxPrefix(acct, app, box string) (string, error) { |
| 127 | return FamilyBoxPrefix(FamilyLogs, acct, app, box) |
| 128 | } |
| 129 | |
| 130 | // HourPrefix is one UTC hour within a box's log subtree. |
| 131 | func HourPrefix(acct, app, box, hour string) (string, error) { |
| 132 | return FamilyHourPrefix(FamilyLogs, acct, app, box, hour) |
| 133 | } |
| 134 | |
| 135 | // ChunkKey names one flushed log chunk: seq is the first record's per-box |
| 136 | // sequence number, zero-padded (16 digits — seq is milliseconds-based, see |
| 137 | // NewSink) so lexical listing order is numeric order, and size is the |
| 138 | // object's own byte length. Size lives IN the key on purpose: |
| 139 | // release.Store's List returns keys only, and retention's byte cap needs |
| 140 | // sizes — embedding them costs a suffix, extending the Store interface |
| 141 | // costs every implementation. Chunks are written once and never appended |
| 142 | // or rewritten (spec §3). |
| 143 | func ChunkKey(acct, app, box, hour string, seq uint64, size int) (string, error) { |
| 144 | return FamilyChunkKey(FamilyLogs, acct, app, box, hour, seq, size) |
| 145 | } |
| 146 | |
| 147 | // ChunkKeySize recovers the byte size a ChunkKey carries, false for any key |
| 148 | // not shaped like one (retention treats unparsable as size-unknown = 0 — |
| 149 | // age still prunes it, the byte cap just cannot count it). |
| 150 | func ChunkKeySize(key string) (int64, bool) { |
| 151 | base := key |
| 152 | if i := strings.LastIndexByte(base, '/'); i >= 0 { |
| 153 | base = base[i+1:] |
| 154 | } |
| 155 | base, ok := strings.CutSuffix(base, ".jsonl.gz") |
| 156 | if !ok { |
| 157 | return 0, false |
| 158 | } |
| 159 | _, sizeStr, ok := strings.Cut(base, "-") |
| 160 | if !ok { |
| 161 | return 0, false |
| 162 | } |
| 163 | n, err := strconv.ParseInt(sizeStr, 10, 64) |
| 164 | if err != nil || n < 0 { |
| 165 | return 0, false |
| 166 | } |
| 167 | return n, true |
| 168 | } |
| 169 | |
| 170 | // EncodeJSONL renders any record shape as gzipped JSONL, one per line — |
| 171 | // the chunk codec every family shares. |
| 172 | func EncodeJSONL[T any](recs []T) ([]byte, error) { |
| 173 | var buf bytes.Buffer |
| 174 | zw := gzip.NewWriter(&buf) |
| 175 | enc := json.NewEncoder(zw) |
| 176 | for _, r := range recs { |
| 177 | if err := enc.Encode(r); err != nil { |
| 178 | return nil, err |
| 179 | } |
| 180 | } |
| 181 | if err := zw.Close(); err != nil { |
| 182 | return nil, err |
| 183 | } |
| 184 | return buf.Bytes(), nil |
| 185 | } |
| 186 | |
| 187 | // DecodeJSONL parses a chunk, skipping malformed lines rather than failing |
| 188 | // the whole object — one corrupt line must not cost the member the rest of |
| 189 | // the hour. Corrupt gzip is an error: that is a broken object, not a line. |
| 190 | func DecodeJSONL[T any](b []byte) ([]T, error) { |
| 191 | zr, err := gzip.NewReader(bytes.NewReader(b)) |
| 192 | if err != nil { |
| 193 | return nil, err |
| 194 | } |
| 195 | defer zr.Close() |
| 196 | var out []T |
| 197 | sc := bufio.NewScanner(zr) |
| 198 | sc.Buffer(nil, 1<<20) // lines are capped well below this at the tee |
| 199 | for sc.Scan() { |
| 200 | var r T |
| 201 | if json.Unmarshal(sc.Bytes(), &r) != nil { |
| 202 | continue |
| 203 | } |
| 204 | out = append(out, r) |
| 205 | } |
| 206 | if err := sc.Err(); err != nil && err != io.EOF { |
| 207 | return out, err |
| 208 | } |
| 209 | return out, nil |
| 210 | } |
| 211 | |
| 212 | // EncodeChunk renders log records: EncodeJSONL over Record. |
| 213 | func EncodeChunk(recs []Record) ([]byte, error) { return EncodeJSONL(recs) } |
| 214 | |
| 215 | // DecodeChunk parses a log chunk: DecodeJSONL over Record. |
| 216 | func DecodeChunk(b []byte) ([]Record, error) { return DecodeJSONL[Record](b) } |
| 217 | |
amadan