Separate Reading Bytes From Interpreting Records

Suppose a text file would contain:

2180,95,true
341,88,false
102,invalid,true

There are at least two different jobs:

1. obtain lines from some source;
2. interpret each line as a domain record.

Do not glue those responsibilities together immediately.

Browser-core parser

Use supplied lines directly:

record MatchRecord(int team, int score, boolean clean) {}

MatchRecord parseLine(String line) {
    String[] parts = line.split(",");

    if (parts.length != 3) {
        throw new IllegalArgumentException("expected 3 fields");
    }

    int team = Integer.parseInt(parts[0].trim());
    int score = Integer.parseInt(parts[1].trim());
    boolean clean = Boolean.parseBoolean(parts[2].trim());

    if (team <= 0 || score < 0) {
        throw new IllegalArgumentException("invalid record values");
    }

    return new MatchRecord(team, score, clean);
}

Parsing has layers

Concept flow

Parsing structured text is a sequence of separate claims

A raw record can satisfy one layer and fail the next.

1. RAW RECORDpreserve the supplied line and identity
   split
2. FIELDScheck shape and expected field count
   convert
3. TYPED VALUESparse numeric and boolean representations
   validate
4. DOMAIN RULESreject values that violate application meaning
   construct
5. RECORDcreate trusted domain state or preserve failure evidence

102,invalid,true has the correct field count but fails numeric conversion.

-4,90,true can convert to integers but violates a domain rule.

Those are different defects.

Boolean parsing has a trap

Boolean.parseBoolean(text) returns true only for case-insensitive "true"; other strings become false instead of throwing an exception.

If your data contract only allows exactly true or false, validate the raw token before trusting a silent false value.

For example, normalize and check membership explicitly.

This is a strong lesson in API contracts: a library method's behavior may not match your application's validation policy.

Parse a batch without losing record identity

Use:

var lines = List.of(
    "2180,95,true",
    "341,88,false",
    "102,invalid,true"
);

Process by line number so a failure report can identify which supplied record failed.

Do not discard the line index/context and print only invalid data.

Evidence

Create a parser for a three-field record. Test:

• one valid ordinary record;
• one boundary-valid record;
• wrong field count;
• invalid numeric field;
• parseable but domain-invalid numeric field;
• invalid boolean token if your contract restricts it.

Preserve the raw line and the stage where it failed.

This is full parsing mastery even without local filesystem access.