EnglishMatrix Sentence Engine: GenerateSentence Reference

The sentence engine consists of three static methods: GenerateSentence, BuildEnglishVerb, and BuildSpanishVerb. Together they take a randomly selected subject, verb, and compatible complement and produce a complete bilingual sentence pair with tense label. The engine is used by all sentence generation modes, both quiz modes, and the tense table mode.

The sentence engine is purely functional — it reads from the static vocabulary lists (subjects, verbsForSentences, complements, verbList, conjugaciones, places, times) and returns strings. It has no side effects and does not mutate any state. The same inputs will always produce the same output.

`SentenceMode` enum

enum SentenceMode { Normal, Question }

Normal

SentenceMode

Produces a declarative sentence in the form Subject + verb phrase + complement + place + time.Example: She eats tacos at home every day. / Ella come tacos en casa todos los días.

Question

SentenceMode

Produces an interrogative sentence by placing the appropriate English auxiliary verb before the subject and appending ?. The Spanish output mirrors the question form using the same conjugated verb preceded by ¿.Example: Does she eat tacos at home every day? / ¿Ella come tacos en casa todos los días?

`GenerateSentence`

static (string english, string spanish) GenerateSentence(
    bool includePlace  = true,
    bool includeTime   = true,
    SentenceMode mode  = SentenceMode.Normal,
    Tense? forcedTense = null
)

GenerateSentence is the main entry point for all output modes. It orchestrates subject, verb, complement, place, and time selection and delegates verb conjugation to BuildEnglishVerb and BuildSpanishVerb.

includePlace

bool

default:"true"

When true, a place expression from the places list is appended to the sentence with 50% probability (a rnd.Next(2) == 0 coin flip). Set to false to always omit place expressions.

includeTime

bool

default:"true"

When true, a time expression from the times list is appended with 50% probability. Set to false to always omit time expressions.

mode

SentenceMode

default:"Normal"

Controls the sentence structure. Normal produces a declarative sentence; Question produces an interrogative sentence with auxiliary inversion.

forcedTense

Tense?

default:"null"

When null, the tense is selected randomly: (Tense)rnd.Next(7), giving each of the 7 tenses equal probability. When set to a specific Tense value, every sentence produced by this call uses that tense.

Return value: A (string english, string spanish) tuple. Each string is prefixed with the Spanish tense label in brackets, e.g. [Presente simple] She eats tacos at home. and [Presente simple] Ella come tacos en casa.

Selection logic (step by step)

Subject: A random entry is picked from subjects (22 entries).
Verb: A random entry is picked from verbsForSentences (79 entries).
Tense: forcedTense is used if set; otherwise (Tense)rnd.Next(7).
Complement: If verb.Accepts != "none", the complements list is filtered to entries whose Type matches verb.Accepts (see Complement compatibility). A random compatible complement is chosen from the filtered list. If the list is empty, no complement is added.
Place: If includePlace is true and rnd.Next(2) == 0, a random entry from places (31 entries) is appended.
Time: If includeTime is true and rnd.Next(2) == 0, a random entry from times (36 entries) is appended.
English verb form: BuildEnglishVerb(subject, verb.English, tense) is called.
Spanish verb form: BuildSpanishVerb(subject, verb.English, tense) is called.
Assembly: For Normal mode: "Subject verbPhrase complement place time." For Question mode: auxiliary inversion is applied (see below).

Question mode — auxiliary inversion

In Question mode, the correct English auxiliary is selected per tense and placed before the subject:

string aux = tense switch
{
    Tense.PresentSimple     => third ? "Does" : "Do",
    Tense.PresentContinuous => third ? "Is" : isNos ? "Are" : "Am",
    Tense.PastSimple        => "Did",
    Tense.PastContinuous    => third ? "Was" : "Were",
    Tense.FutureSimple      => "Will",
    Tense.PresentPerfect    => third ? "Has" : "Have",
    Tense.Conditional       => "Would",
    _ => "Do"
};

For compound tenses (PresentContinuous, PastContinuous, PresentPerfect), where the auxiliary is already embedded in engVerb, the engine strips the auxiliary from engVerb and uses only the participle/gerund component as the question verb:

bool verbAlreadyFormed = tense == Tense.PresentContinuous
                      || tense == Tense.PastContinuous
                      || tense == Tense.PresentPerfect;

string qVerb = verbAlreadyFormed
    ? engVerb.Substring(engVerb.IndexOf(' ') + 1)
    : verb.English;

The resulting question is assembled as: "Aux Subject qVerb complement place time?"

`BuildEnglishVerb`

static string BuildEnglishVerb(string subject, string verb, Tense tense)

Produces the correctly inflected English verb phrase for a given subject and tense. Does not include the subject itself in its return value.

subject

string

required

The English subject string (e.g., "I", "She", "The teacher"). Used to determine person/number for auxiliary selection.

verb

string

required

The bare English infinitive (e.g., "eat", "run"). Irregular forms are looked up via GetPast(), GetParticiple(), and GetGerund().

tense

Tense

required

The tense to build. One of the 7 Tense enum values.

Returns: The conjugated verb phrase string (without subject), e.g. "eats", "is eating", "ate", "was eating", "will eat", "has eaten", "would eat".

Tense-by-tense formation

Tense	Return value	Notes
`PresentSimple`	`ConjugateEnglishPresent(subject, verb)`	Applies `-s`/`-es`/`-ies` for 3rd-person singular; bare verb for all others
`PresentContinuous`	`"{be} {GetGerund(verb)}"`	`be` = `am`/`is`/`are` based on subject
`PastSimple`	`GetPast(verb)`	Looks up `VerbEntry.Past`; falls back to `verb + "ed"` if not found
`PastContinuous`	`"{was} {GetGerund(verb)}"`	`was` for I/she/he/it and 3rd-person; `were` for we/they/you
`FutureSimple`	`"will {verb}"`	Uniform across all subjects
`PresentPerfect`	`"{has/have} {GetParticiple(verb)}"`	`has` for 3rd-person singular; `have` for all others
`Conditional`	`"would {verb}"`	Uniform across all subjects

Auxiliary selection for `be`

string be;
if (third)
    be = "is";
else if (subjLower == "we" || subjLower == "they" || subjLower == "you")
    be = "are";
else
    be = "am";

Auxiliary selection for `was`/`were`

string was;
if (third || subjLower == "i" || subjLower == "he" || subjLower == "she" || subjLower == "it")
    was = "was";
else
    was = "were";

Irregular form lookup helpers

GetPast(verb) — returns VerbEntry.Past or verb + "ed" as fallback
GetParticiple(verb) — returns VerbEntry.Participle or verb + "ed" as fallback
GetGerund(verb) — returns VerbEntry.Gerund or verb + "ing" as fallback

`BuildSpanishVerb`

static string BuildSpanishVerb(string subject, string verb, Tense tense)

Produces the conjugated Spanish verb phrase for a given English subject and verb. Uses the conjugaciones dictionary as the primary source and falls back to programmatic conjugation when an entry is missing.

subject

string

required

The English subject string. Used to determine person mapping: "I" → yo (first singular), "we" → nosotros (first plural), all others → él/ella (third singular).

verb

string

required

The bare English infinitive, used as the key to look up conjugaciones.

tense

Tense

required

The tense to build.

Returns: The conjugated Spanish verb phrase string, e.g. "como", "estoy comiendo", "comí", "estaba comiendo", "voy a comer", "he comido", "comería".

Person mapping

The engine uses only two boolean flags, derived directly from the English subject string:

bool isYo  = subject.ToLower() == "i";
bool isNos = subject.ToLower() == "we";

English subject(s)	Mapped person	Fields used
`"I"`	yo (1st singular)	`YoPres`, `YoPast`
`"We"`	nosotros (1st plural)	`NosPres`, `NosPast`
All others — `"She"`, `"He"`, `"They"`, `"You"`, `"My mom"`, `"The teacher"`, etc.	él/ella (3rd singular)	`ElPres`, `ElPast`

"They" and "You" are not isNos — they fall through to the default ElPres/ElPast branch alongside all other non-"I", non-"We" subjects.

Tense-by-tense formation (when verb is in `conjugaciones`)

Tense	yo	él/ella	nosotros
`PresentSimple`	`c.YoPres`	`c.ElPres`	`c.NosPres`
`PresentContinuous`	`"estoy {gerund}"`	`"está {gerund}"`	`"estamos {gerund}"`
`PastSimple`	`c.YoPast`	`c.ElPast`	`c.NosPast`
`PastContinuous`	`"estaba {gerund}"`	`"estaba {gerund}"`	`"estábamos {gerund}"`
`FutureSimple`	`"voy a {inf}"`	`"va a {inf}"`	`"vamos a {inf}"`
`PresentPerfect`	`"he {participle}"`	`"ha {participle}"`	`"hemos {participle}"`
`Conditional`	`BuildConditional(inf, isYo, isNos, conditionalStems)`	—	—

Where gerund = GetSpanishGerund(verb), inf = GetSpanishInfinitive(verb), and participle = GetSpanishPastParticiple(verb).

Spanish gerund computation (`GetSpanishGerund`)

static string GetSpanishGerund(string verb)
{
    string inf = GetSpanishInfinitive(verb);
    if (inf.EndsWith("ar")) return inf.Substring(0, inf.Length - 2) + "ando";
    if (inf.EndsWith("er")) return inf.Substring(0, inf.Length - 2) + "iendo";
    if (inf.EndsWith("ir")) return inf.Substring(0, inf.Length - 2) + "iendo";
    
    if (inf == "ir")    return "yendo";
    if (inf == "leer")  return "leyendo";
    if (inf == "creer") return "creyendo";
    if (inf == "oír")   return "oyendo";
    
    return inf;
}

Spanish past participle computation (`GetSpanishPastParticiple`)

Applies regular -ado/-ido endings and handles common irregular participles:

static string GetSpanishPastParticiple(string verb)
{
    string inf = GetSpanishInfinitive(verb);
    if (inf.EndsWith("ar"))              return inf.Substring(0, inf.Length - 2) + "ado";
    if (inf.EndsWith("er") || inf.EndsWith("ir"))
                                         return inf.Substring(0, inf.Length - 2) + "ido";
    
    if (inf == "abrir")   return "abierto";
    if (inf == "cubrir")  return "cubierto";
    if (inf == "decir")   return "dicho";
    if (inf == "escribir")return "escrito";
    if (inf == "hacer")   return "hecho";
    if (inf == "morir")   return "muerto";
    if (inf == "poner")   return "puesto";
    if (inf == "romper")  return "roto";
    if (inf == "ver")     return "visto";
    if (inf == "volver")  return "vuelto";
    
    return inf;
}

Fallback for missing conjugations

When a verb is not found in conjugaciones, BuildSpanishVerb constructs approximate Spanish forms directly from the Spanish infinitive’s ending:

if (!conjugaciones.TryGetValue(verb.ToLower(), out var c))
{
    string regularPresent = isYo ? inf.Replace("ar","o").Replace("er","o").Replace("ir","o")
                         : isNos ? inf.Replace("ar","amos").Replace("er","emos").Replace("ir","imos")
                         : inf.Replace("ar","a").Replace("er","e").Replace("ir","e");

    string regularPast = isYo ? inf.Replace("ar","é").Replace("er","í").Replace("ir","í")
                       : isNos ? inf.Replace("ar","amos").Replace("er","imos").Replace("ir","imos")
                       : inf.Replace("ar","ó").Replace("er","ió").Replace("ir","ió");
    // ... returns tense-appropriate form
}

This fallback produces correct output for regular verbs but may be inaccurate for irregular verbs that lack a conjugaciones entry. 35 verbs in verbList are currently affected — see Verb coverage gap for the full list.

Conditional stem changes

The conditionalStems dictionary inside BuildSpanishVerb maps Spanish infinitives to their irregular conditional stems before the -ía/-íamos ending is applied:

var conditionalStems = new Dictionary<string, string>
{
    { "tener", "tendr" }, { "venir", "vendr" }, { "poder", "podr" },
    { "poner", "pondr" }, { "salir", "saldr" }, { "hacer", "har"  },
    { "decir", "dir"   }, { "querer","querr"  }, { "saber","sabr"  }
};

BuildConditional applies the irregular stem if found, then appends the appropriate ending:

static string BuildConditional(string inf, bool isYo, bool isNos,
    Dictionary<string, string> stems)
{
    string stem = inf;
    foreach (var kvp in stems)
        if (inf.Contains(kvp.Key)) { stem = inf.Replace(kvp.Key, kvp.Value); break; }
    
    if (stem.EndsWith("ar"))
        return isYo ? stem.Replace("ar","aría")
             : isNos ? stem.Replace("ar","aríamos")
             : stem.Replace("ar","aría");
    if (stem.EndsWith("er"))
        return isYo ? stem.Replace("er","ería")
             : isNos ? stem.Replace("er","eríamos")
             : stem.Replace("er","ería");
    if (stem.EndsWith("ir"))
        return isYo ? stem.Replace("ir","iría")
             : isNos ? stem.Replace("ir","iríamos")
             : stem.Replace("ir","iría");
    
    return isYo ? $"{inf}ía" : isNos ? $"{inf}íamos" : $"{inf}ía";
}

Get Started

Modes & Features

Reference

EnglishMatrix Sentence Engine: GenerateSentence Reference

`SentenceMode` enum

`GenerateSentence`

Selection logic (step by step)

Question mode — auxiliary inversion

`BuildEnglishVerb`

Tense-by-tense formation

Auxiliary selection for `be`

Auxiliary selection for `was`/`were`

Irregular form lookup helpers

`BuildSpanishVerb`

Person mapping

Tense-by-tense formation (when verb is in `conjugaciones`)

Spanish gerund computation (`GetSpanishGerund`)

Spanish past participle computation (`GetSpanishPastParticiple`)

Fallback for missing conjugations

Conditional stem changes

Build docs developers (and LLMs) love

Get Started

Modes & Features

Reference

Documentation Index

​SentenceMode enum

​GenerateSentence

​Selection logic (step by step)

​Question mode — auxiliary inversion

​BuildEnglishVerb

​Tense-by-tense formation

​Auxiliary selection for be

​Auxiliary selection for was/were

​Irregular form lookup helpers

​BuildSpanishVerb

​Person mapping

​Tense-by-tense formation (when verb is in conjugaciones)

​Spanish gerund computation (GetSpanishGerund)

​Spanish past participle computation (GetSpanishPastParticiple)

​Fallback for missing conjugations

​Conditional stem changes

Build docs developers (and LLMs) love

`SentenceMode` enum

`GenerateSentence`

Selection logic (step by step)

Question mode — auxiliary inversion

`BuildEnglishVerb`

Tense-by-tense formation

Auxiliary selection for `be`

Auxiliary selection for `was`/`were`

Irregular form lookup helpers

`BuildSpanishVerb`

Person mapping

Tense-by-tense formation (when verb is in `conjugaciones`)

Spanish gerund computation (`GetSpanishGerund`)

Spanish past participle computation (`GetSpanishPastParticiple`)

Fallback for missing conjugations

Conditional stem changes